[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Next Issue
Volume 21, April-2
Previous Issue
Volume 21, March-2
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
7.3
3.4
Journals
Sensors
Volume 21
Issue 7
sensors-logo

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Sensors, Volume 21, Issue 7 (April-1 2021) – 319 articles

Cover Story (view full-size image): In smart cities, spatio-temporal interpolation provides a fine-grained understanding of local phenomena, such as weather, air quality, or traffic data, offering estimates of observations in unobserved locations and time slots. However, with the ever-increasing sensor data, the data transmission and processing requirements of the predominantly centralized architectures have become unfeasible. To address this scaling problem, we propose EDISON: algorithms for distributed learning and inference, and an edge-native architecture for distributing the data and the computations between device, edge and cloud layers. The results show that EDISON provides an improvement over alternative approaches, reaching at best a 10% smaller RMSE than a global interpolation, and a 6% smaller RMSE than a baseline-distributed approach. View this paper.
  • Issues are regarded as officially published after their release is announced to the table of contents alert mailing list.
  • You may sign up for e-mail alerts to receive table of contents of newly released issues.
  • PDF is the official format for papers published in both, html and pdf forms. To view the papers in pdf format, click on the "PDF Full-text" link, and use the free Adobe Reader to open them.
Order results
Result details
Section
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
15 pages, 1898 KiB  
Article
A Comprehensive Comparison and Validation of Published Methods to Detect Turn Switch during Alpine Skiing
by Aaron Martínez, Cory Snyder, Stephanie R. Moore and Thomas Stöggl
Sensors 2021, 21(7), 2573; https://doi.org/10.3390/s21072573 - 6 Apr 2021
Cited by 6 | Viewed by 2957
Abstract
The instant of turn switch (TS) in alpine skiing has been assessed with a variety of sensors and TS concepts. Despite many published methodologies, it is unclear which is best or how comparable they are. This study aimed to facilitate the process of [...] Read more.
The instant of turn switch (TS) in alpine skiing has been assessed with a variety of sensors and TS concepts. Despite many published methodologies, it is unclear which is best or how comparable they are. This study aimed to facilitate the process of choosing a TS method by evaluating the accuracy and precision of the methodologies previously used in literature and to assess the influence of the sensor type. Optoelectronic motion capture, inertial measurement units, pressure insoles, portable force plates, and electromyography signals were recorded during indoor treadmill skiing. All TS methodologies were replicated as stated in their respective publications. The method proposed by Supej assessed with optoelectronic motion capture was used as a comparison reference. TS time differences between the reference and each methodology were used to assess accuracy and precision. All the methods analyzed showed an accuracy within 0.25 s, and ten of them within 0.05 s. The precision ranged from ~0.10 s to ~0.60 s. The TS methodologies with the best performance (accuracy and precision) were Klous Video, Spörri PI (pressure insoles), Martinez CTD (connected boot), and Yamagiwa IMU (inertial measurement unit). In the future, the specific TS methodology should be chosen with respect to sensor selection, performance, and intended purpose. Full article
(This article belongs to the Section Physical Sensors)
►▼ Show Figures

Figure 1

Figure 1
<p>Illustration of the experiment configuration in the indoor skiing carpet.</p> Full article ">Figure 2
<p>Graphical illustration and equations to clarify the interpretation of the data presented in the results timeline. <span class="html-italic">i</span>, participant index; n, number of participants per method; <span class="html-italic">SD</span> , standard deviation; CI<sub>lower</sub>, lower confidence interval limit (percentile 2.5); CI<sub>upper</sub>, upper confidence interval limit (percentile 97.5).</p> Full article ">Figure 3
<p>Timeline depicting the accuracy and precision of the turn switch detection methodology. The left side (<b>a</b>) figure shows the accuracy (black dot), standard deviation of the accuracy (whiskers) and precision (shaded area) of every method calculated with all turns pooled together. The right side (<b>b</b>) graph separates the data by turn size, long turns (red) and short turns (blue).</p> Full article ">
4 pages, 172 KiB  
Editorial
Biomedical Photoacoustic Imaging and Sensing Using Affordable Resources
by Mithun Kuniyil Ajith Singh and Wenfeng Xia
Sensors 2021, 21(7), 2572; https://doi.org/10.3390/s21072572 - 6 Apr 2021
Cited by 4 | Viewed by 2675
Abstract
The photoacoustic (PA) effect, also called the optoacoustic effect, was discovered in the 1880s by Alexander Graham Bell and has been utilized for biomedical imaging and sensing applications since the early 1990s [...] Full article
(This article belongs to the Special Issue Biomedical Photoacoustic Imaging and Sensing Using Affordable Resources)
28 pages, 3214 KiB  
Article
Semantic Interoperability between IEC 61850 and oneM2M for IoT-Enabled Smart Grids
by Salvatore Cavalieri
Sensors 2021, 21(7), 2571; https://doi.org/10.3390/s21072571 - 6 Apr 2021
Cited by 14 | Viewed by 3922
Abstract
In the era of Industry 4.0, pervasive adoption of communication technologies based on the Internet of Things represents a very strong requirement in several domains. In the smart grid domain, there is the need to overcome one of the main limitations of the [...] Read more.
In the era of Industry 4.0, pervasive adoption of communication technologies based on the Internet of Things represents a very strong requirement in several domains. In the smart grid domain, there is the need to overcome one of the main limitations of the current electric grid, allowing the use of heterogeneous devices capable of measuring, monitoring and exchanging information about grid components. For this reason, current literature often presents research activities about enabling internet of things (IoT) in smart grids; in particular, several proposals aim to realize interworking between IoT and smart grid communication standards, allowing exchange of information between IoT devices and the electrical grid components. Semantic interoperability should be achieved in an interworking solution in order to provide a common meaning of the data exchanged by heterogeneous devices, even if they belong to different domains. Until now, semantic interoperability remains an open challenge in the smart grid field. The paper aims to propose a novel solution of interworking between two of the most used communication systems in smart grids and IoT domains, i.e., IEC 61850 and oneM2M, respectively. A semantic interoperability solution is also proposed to be used in the interworking scheme here presented. Full article
(This article belongs to the Special Issue Industrial Internet of Things in the Industry 4.0: New Researches, Applications and Challenges)
►▼ Show Figures

Figure 1

Figure 1
<p>IEC 61850 Information Model.</p> Full article ">Figure 2
<p>IEC61950-SAREF4ENER Ontology.</p> Full article ">Figure 3
<p>oneM2M Infrastructure and Field Domains.</p> Full article ">Figure 4
<p>Hierarchical structure of a oneM2M resource.</p> Full article ">Figure 5
<p>Common Service Entity (CSE) resource tree.</p> Full article ">Figure 6
<p>Interworking Proxy application Entity (IPE).</p> Full article ">Figure 7
<p>Classes and Properties of oneM2M Base Ontology.</p> Full article ">Figure 8
<p>CSE structure in oneM2M-based Interworking.</p> Full article ">Figure 9
<p>Common Ontology between oneM2M Base Ontology and SAREF-based Washing Machine Ontology.</p> Full article ">Figure 10
<p>XSD file representing the ABC: ABC_WM class as a &lt;flexContainer&gt; oneM2M resource.</p> Full article ">Figure 11
<p>CSE structure considering the Common Ontology of <a href="#sensors-21-02571-f009" class="html-fig">Figure 9</a>.</p> Full article ">Figure 12
<p>Common Ontology between IEC61850-SARF4ENER and oneM2M Base Ontology.</p> Full article ">Figure 13
<p>Interworking Architecture based on IPE and Common Ontology.</p> Full article ">Figure 14
<p>XSD file representing the iec61850s4ener: LogicalNode class as a &lt;flexContainer&gt; oneM2M resource.</p> Full article ">Figure 15
<p>Operations performed by IPE at start-up.</p> Full article ">Figure 16
<p>Example of the oneM2M resources created by IPE in the Registrar CSE at Start-up.</p> Full article ">Figure 17
<p>Retrieving information produced by IEC 61850 Server.</p> Full article ">Figure 18
<p>Updating information to IEC 61850 Server.</p> Full article ">Figure 19
<p>Invoking services running on IEC 61850 Services from oneM2M domain.</p> Full article ">
15 pages, 20227 KiB  
Article
Response of Transitional Mixtures Retaining Memory of In-Situ Overburden Pressure Monitored Using Electromagnetic and Piezo Crystal Sensors
by Sang Yeob Kim, Jong-Sub Lee and Junghee Park
Sensors 2021, 21(7), 2570; https://doi.org/10.3390/s21072570 - 6 Apr 2021
Cited by 1 | Viewed by 2571
Abstract
The major and minor components in granular soil materials determine their properties and behavior. This study explores the transitional behavior within threshold fines fraction of soil mixtures based on the data from the literature and experiments. From the literature survey, the void ratio, [...] Read more.
The major and minor components in granular soil materials determine their properties and behavior. This study explores the transitional behavior within threshold fines fraction of soil mixtures based on the data from the literature and experiments. From the literature survey, the void ratio, shear wave velocity, compression index, and friction angle capture the transitional turning point between the low and data-adjusted high threshold fines fractions. In particular, there is a dramatic change in hydraulic conductivity below the low threshold fines fraction that highlights the critical role of small amounts of fines in the fluid flow (e.g., clogging). From an experimental study, the engineering properties of natural soil samples identified using deformation and elastic wave sensors show transitional trends within the Revised Soil Classification System framework. The evolution of compressibility and shear wave velocity indicate that either coarse, fine, or both particles are likely to contribute to large and small strain stiffnesses when the effective stress is below 400 kPa. Thereafter, both engineering properties indicate that the soil sample retains a memory of in-situ overburden pressure when the effective stress is around 400 kPa. There is a critical role of fines that are slightly higher than low threshold fines fraction on engineering properties that promote the application of Revised Soil Classification System RSCS to natural soils. Full article
(This article belongs to the Special Issue Emerging Characterization of Geomaterials Using Advanced Geo-Sensors)
►▼ Show Figures

Figure 1

Figure 1
<p>Porosity versus fines fraction. Two red dotted lines indicate low and data-adjusted high-threshold fines fractions computed using Equations (2) and (4). Input parameters used for estimation of two boundaries are <span class="html-italic">e<sub>C</sub><sup>max</sup></span> = 0.805, <span class="html-italic">e<sub>C</sub><sup>min</sup></span> = 0.533, <span class="html-italic">e<sub>F</sub><sup>max</sup></span> = 0.949; <span class="html-italic">e<sub>F</sub><sup>min</sup></span> = 0.689, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>L</sup></span> = 21.5%, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>H</sup></span>* = 54.4%. Note: Data extracted from Vallejo [<a href="#B30-sensors-21-02570" class="html-bibr">30</a>].</p> Full article ">Figure 2
<p>Shear wave velocity versus fines fraction. Two red dotted lines indicate low and data-adjusted high-threshold fines fractions computed using Equations (2) and (4). Input parameters used for estimation of two boundaries are <span class="html-italic">e<sub>C</sub><sup>max</sup></span> = 0.85, <span class="html-italic">e<sub>C</sub><sup>min</sup></span> = 0.59, <span class="html-italic">e<sub>F</sub><sup>max</sup></span> = 0.88; <span class="html-italic">e<sub>F</sub><sup>min</sup></span> = 0.56, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>L</sup></span> = 23.9%, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>H</sup></span>* = 41.5%. Note: Data extracted from Lee et al. [<a href="#B21-sensors-21-02570" class="html-bibr">21</a>].</p> Full article ">Figure 3
<p>Compression index versus fines fraction. Two red dotted lines indicate low and data-adjusted high-threshold fines fractions computed using Equations (2) and (4). Input parameters used for estimation of two boundaries are <span class="html-italic">e<sub>C</sub><sup>max</sup></span> = 0.912, <span class="html-italic">e<sub>C</sub><sup>min</sup></span> = 0.584, <span class="html-italic">e<sub>F</sub><sup>max</sup></span> = 1.47; <span class="html-italic">e<sub>F</sub><sup>min</sup></span> = 0.53, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>L</sup></span> = 19.1%, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>H</sup></span>* = 43.6%. Note: Data extracted from Simpson and Evans [<a href="#B31-sensors-21-02570" class="html-bibr">31</a>].</p> Full article ">Figure 4
<p>Friction angle versus fines fraction. Two red dotted lines indicate low and data-adjusted high-threshold fines fractions computed using Equations (2) and (4). Input parameters used for estimation of two boundaries are <span class="html-italic">e<sub>C</sub><sup>max</sup></span> = 0.737, <span class="html-italic">e<sub>C</sub><sup>min</sup></span> = 0.548, <span class="html-italic">e<sub>F</sub><sup>max</sup></span> = 0.879; <span class="html-italic">e<sub>F</sub><sup>min</sup></span> = 0.56, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>L</sup></span> = 22.6%, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>H</sup></span>* = 54.2%. Note: Data extracted from Ueda et al. [<a href="#B32-sensors-21-02570" class="html-bibr">32</a>].</p> Full article ">Figure 5
<p>Hydraulic conductivity versus fines fraction. Red dotted lines indicate low threshold fines fraction computed using Equation (2). Input parameters used for estimation of two boundaries are <span class="html-italic">e<sub>C</sub><sup>min</sup></span> = 0.45 and <span class="html-italic">e<sub>F</sub><sup>max</sup></span> = 1.51, and <span class="html-italic">F<sub>th</sub></span>|<span class="html-italic"><sup>L</sup></span> = 26.5%. Note: Data extracted from Zhang et al. [<a href="#B33-sensors-21-02570" class="html-bibr">33</a>].</p> Full article ">Figure 6
<p>Particle size distribution of Savannah River sand (boring number: HPC-1, Sampling depth: 19.20 m–19.35 m).</p> Full article ">Figure 7
<p>Floating oedometer cell with bender elements and peripheral devices for measurement of shear waves (note: BE denotes bender element).</p> Full article ">Figure 8
<p>Load-deformation response for original field sample under zero-later strain conditions.</p> Full article ">Figure 9
<p>Compaction curves for three tested specimens in view of void ratio <span class="html-italic">e</span> against vertical effective stress <span class="html-italic">σ</span>′<span class="html-italic"><sub>v</sub></span>.</p> Full article ">Figure 10
<p>Compressibility for Savannah River sand estimated at consecutive loading steps versus vertical effective stress (note: Sampling depth is 19.20 m–19.35 m).</p> Full article ">Figure 11
<p>Shear wave signals for Savannah River sands during K<sub>o</sub> loading and unloading: (<b>a</b>) Original field sample; (<b>b</b>) Fine-grained soil passed through No. 100 sieve; (<b>c</b>) Coarse-grained soil retained on No. 100 sieve.</p> Full article ">Figure 12
<p>Shear wave velocity for Savannah River sands versus vertical effective stress (note: Sampling depth is 19.20 m–19.35 m).</p> Full article ">Figure 13
<p>Conceptual drawing of changes in mechanical response and fluid flow in terms of low and data-adjusted high-threshold fines fractions.</p> Full article ">
23 pages, 6255 KiB  
Article
Complete Automation of an Energy Consumption Reduction Strategy from a Water Treatment and Distribution Facility, Inside an Industrial Internet of Things-Compliant Proactive Historian Application
by Andrei Nicolae, Adrian Korodi and Ioan Silea
Sensors 2021, 21(7), 2569; https://doi.org/10.3390/s21072569 - 6 Apr 2021
Cited by 7 | Viewed by 2739
Abstract
The Industrial Internet of Things and Industry 4.0 paradigms are steering the industrial landscape towards better connected entities, superior interoperability and information exchange, which lays the basis for developing more intelligent solutions that are already starting to bring numerous benefits. The current research [...] Read more.
The Industrial Internet of Things and Industry 4.0 paradigms are steering the industrial landscape towards better connected entities, superior interoperability and information exchange, which lays the basis for developing more intelligent solutions that are already starting to bring numerous benefits. The current research aligns to this course, in an attempt to build an automated and autonomous software tool, capable of reducing the energy consumption of a water treatment and distribution facility, by optimizing the water sources usage. Based on several previous researches, the present paper details both the complete automation of the optimizing strategy inside a proactive historian application and the tests executed with the finished solution. Possessing the abilities to directly influence the monitored system in a non-invasive manner, and to link all the sequences of the algorithm automatically, the solution is now ready for long-term functioning without any external interference. Full article
(This article belongs to the Special Issue Sensors and Real Time Systems for IIoT)
►▼ Show Figures

Figure 1

Figure 1
<p>Simplified overview of the stages taken by the drinking water before entering the consumer network.</p> Full article ">Figure 2
<p>Example of the minimum objects that must be defined, inside the Historian Process Editor, for the currently used process, in order to execute the optimizing strategy.</p> Full article ">Figure 3
<p>Water source characteristics that must have tags associated, inside the Historian Process Editor, for the currently used process, in order to execute the optimizing strategy.</p> Full article ">Figure 4
<p>Optimizing objectives choice inside the proactive Historian application.</p> Full article ">Figure 5
<p>The implemented algorithm which determines <span class="html-italic">q<sub>f</sub></span> from the results of the first level dependencies identification algorithm.</p> Full article ">Figure 6
<p>A high-level summary of the implemented algorithm for dividing the total requested water flow into specific flows for each source, in an optimum way for reducing energy consumption.</p> Full article ">Figure 7
<p>Detailed perspective of the implemented algorithm for dividing the total requested water flow into specific flows for each source, in an optimum way for reducing energy consumption, (<b>a</b>) right side of the diagram, (<b>b</b>) left side of the diagram.</p> Full article ">Figure 7 Cont.
<p>Detailed perspective of the implemented algorithm for dividing the total requested water flow into specific flows for each source, in an optimum way for reducing energy consumption, (<b>a</b>) right side of the diagram, (<b>b</b>) left side of the diagram.</p> Full article ">Figure 8
<p>Example of displaying, in the Historian GUI, the results of running the optimization solution.</p> Full article ">
17 pages, 41705 KiB  
Article
Common-Mode Voltage Reduction in Capacitive Sensing of Biosignal Using Capacitive Grounding and DRL Electrode
by Tadeas Bednar, Branko Babusiak, Michal Labuda, Milan Smetana and Stefan Borik
Sensors 2021, 21(7), 2568; https://doi.org/10.3390/s21072568 - 6 Apr 2021
Cited by 5 | Viewed by 3749
Abstract
A capacitive measurement of the biosignals is a very comfortable and unobtrusive way suitable for long-term and wearable monitoring of health conditions. This type of sensing is very susceptible to noise from the surroundings. One of the main noise sources is power-line noise, [...] Read more.
A capacitive measurement of the biosignals is a very comfortable and unobtrusive way suitable for long-term and wearable monitoring of health conditions. This type of sensing is very susceptible to noise from the surroundings. One of the main noise sources is power-line noise, which acts as a common-mode voltage at the input terminals of the acquisition unit. The origin and methods of noise reduction are described on electric models. Two methods of noise removal are modeled and experimentally verified in the paper. The first method uses a passive capacitive grounding electrode, and the second uses an active capacitive Driven Right Leg (DRL) electrode. The effect of grounding electrode size on noise suppression is experimentally investigated. The increasing electrode area reduces power-line noise: the power of power-line frequency within the measured signal is 70.96 dB, 59.13 dB, and 43.44 dB for a grounding electrode area of 1650 cm2, 3300 cm2, and 4950 cm2, respectively. The capacitive DRL electrode shows better efficiency in common-mode noise rejection than the grounding electrode. When using an electrode area of 1650 cm2, the DRL achieved 46.3 dB better attenuation than the grounding electrode at power-line frequency. In contrast to the grounding electrode, the DRL electrode reduces a capacitive measurement system’s financial costs due to the smaller electrode area made of the costly conductive textile. Full article
(This article belongs to the Special Issue Sensors, Circuit and System for Biomedical Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Model of the active electrode.</p> Full article ">Figure 2
<p>Transfer characteristics according to different values of <span class="html-italic">R</span><sub>B</sub>.</p> Full article ">Figure 3
<p>Schematic (<b>a</b>) and design (<b>b</b>) of the constructed active electrode. Reprinted with permission from ref. [<a href="#B16-sensors-21-02568" class="html-bibr">16</a>]. Copyright 2019 Elsevier B.V.</p> Full article ">Figure 4
<p>Measured transfer characteristics.</p> Full article ">Figure 5
<p>Electrical model depicting the power-line interference.</p> Full article ">Figure 6
<p>Schematic circuit (<b>a</b>) and equivalent circuit for a common-mode voltage (<b>b</b>).</p> Full article ">Figure 7
<p>Common-mode voltage rejection for a system with noise suppression electrode (NSE) electrode connected to the system ground.</p> Full article ">Figure 8
<p>Equivalent schematic of an interference model for the system using a Driven Right Leg (DRL) circuit.</p> Full article ">Figure 9
<p>Common-mode voltage rejection for the system using a DRL circuit.</p> Full article ">Figure 10
<p>The comparison of common-mode rejection depending on different DRL gain.</p> Full article ">Figure 11
<p>Block scheme of developed electrocardiography (ECG) system.</p> Full article ">Figure 12
<p>Placement of electrodes on the bedsheet (<b>a</b>), and a model of coupling capacitances between the subject and electrodes (<b>b</b>).</p> Full article ">Figure 13
<p>Schematic of the simulation model.</p> Full article ">Figure 14
<p>The ECG signal waveform (<b>a</b>) and the power-line voltage waveform (<b>b</b>) used in the simulation.</p> Full article ">Figure 15
<p>The output signals from simulation while using the same coupling capacitances (<span class="html-italic">C</span><sub>e1</sub> = <span class="html-italic">C</span><sub>e2</sub> = 150 pF) at the input in two scenarios: The NSE connected to the ground (solid blue), and NSE connected to the output of DRL (dashed red).</p> Full article ">Figure 16
<p>The output signals from simulation while using different coupling capacitances (<span class="html-italic">C</span><sub>e1</sub> = 150 pF, <span class="html-italic">C</span><sub>e2</sub> = 75 pF) at the input in two scenarios: The NSE connected to the ground (solid blue), and NSE connected to the output of DRL (solid red).</p> Full article ">Figure 17
<p>Capacitive ECG with gradually connected grounding electrodes.</p> Full article ">Figure 18
<p>Filtered ECG data with gradually connected grounding electrodes—no grounding electrode (<b>a</b>), one grounding electrode (<b>b</b>), two grounding electrodes (<b>c</b>), and three grounding electrodes (<b>d</b>).</p> Full article ">Figure 19
<p>Raw ECG signal (<b>a</b>) and filtered signal (<b>b</b>) acquired by the capacitive system using the DRL electrode.</p> Full article ">Figure 20
<p>Power spectral densities of capacitive ECG signal using grounding (<b>blue</b>) and DRL (<b>red</b>) electrode.</p> Full article ">
14 pages, 61494 KiB  
Article
Quantifying Hole-Edge Crack of Bolt Joints by Using an Embedding Triangle Eddy Current Sensing Film
by Shilei Fan, Junyan Yi, Hu Sun and Fenglin Yun
Sensors 2021, 21(7), 2567; https://doi.org/10.3390/s21072567 - 6 Apr 2021
Cited by 5 | Viewed by 3163
Abstract
Hole-edge crack quantification of bolt joints is critical for monitoring and estimating structural integrity of aircraft. The paper proposes a new triangle eddy current sensor array for the purpose of increasing the level of quantifying hole-edge crack parameters, especially, the crack angle. The [...] Read more.
Hole-edge crack quantification of bolt joints is critical for monitoring and estimating structural integrity of aircraft. The paper proposes a new triangle eddy current sensor array for the purpose of increasing the level of quantifying hole-edge crack parameters, especially, the crack angle. The new senor array consists of triangular coils instead of planar rectangular coils. The configuration of the novel sensor array, including the excitation current directions and the excitation winding shape, is optimized by simulation. The ability of the proposed sensing film to identify the crack parameters has been verified by finite element simulations and experiments. Results shows that triangular coils with same current directions in circumferentially adjacent coils and opposite current directions in axially adjacent coils achieve better performance in sensor linearity and resolution compared to rectangular coils. In addition, it has also been proved that the sensing film has a good potential to identify the crack depth and length. Full article
(This article belongs to the Special Issue Smart Sensors for Structural Health Monitoring and Nondestructive Evaluation)
►▼ Show Figures

Figure 1

Figure 1
<p>An eddy current (EC) sensing film-based intelligent bolt.</p> Full article ">Figure 2
<p>The simplified triangular sensing coil array configuration.</p> Full article ">Figure 3
<p>The simulation model.</p> Full article ">Figure 4
<p>Four exciting configurations.</p> Full article ">Figure 5
<p>Simulation results for determining the best exciting configuration. (<b>a</b>) Voltage variations of the S1 and S2 when the crack is set at C5R1. (<b>b</b>) Voltage variations of S2 when the crack propagates in the radial direction at C5.</p> Full article ">Figure 6
<p>Current density distribution.</p> Full article ">Figure 7
<p>EC vector and cloud picture when the crack is at C5R1.</p> Full article ">Figure 8
<p>Simulation results of the crack angle detection. (<b>a</b>) Model 1 (rectangular exciting coils); (<b>b</b>) Model 2 (triangular exciting coils).</p> Full article ">Figure 9
<p>A schematic diagram of the inductive current directions in one triangular turn in Model 1.</p> Full article ">Figure 10
<p>The side view of EC in Model 1.</p> Full article ">Figure 11
<p>The difference voltage of the S1 and S2 versus the crack circumferential position.</p> Full article ">Figure 12
<p>Simulation results for the second layer. (<b>a</b>) Voltage variation of each coil; (<b>b</b>) Difference value between adjacent coils.</p> Full article ">Figure 13
<p>The induced voltage variation of S1 and S2 when the crack propagates along the radial direction at each circumferential position. (<b>a</b>) S1; (<b>b</b>) S2.</p> Full article ">Figure 14
<p>Voltage variation versus axial length.</p> Full article ">Figure 15
<p>Experimental setup and sensor film. (<b>a</b>) Signal generator; (<b>b</b>) intelligent bolt; (<b>c</b>) signal receiver; (<b>d</b>) signal displayer; (<b>e</b>) eddy current sensing film.</p> Full article ">Figure 16
<p>Crack angle detection results. (<b>a</b>) Voltage variation of the S2–S4 versus crack angle; (<b>b</b>) Voltage difference of S3 (S2) and S4 (S3).</p> Full article ">Figure 17
<p>Experimental results when the crack grows in the radial direction at two circumferential positions. (<b>a</b>) Voltage variation when the crack position is at the middle of the S2; (<b>b</b>) voltage variation when the crack position is in the middle of the S2 and S3.</p> Full article ">Figure 18
<p>Experimental results as the crack grows in the axial direction.</p> Full article ">
13 pages, 1592 KiB  
Article
Gas Damping in Capacitive MEMS Transducers in the Free Molecular Flow Regime
by Boris A. Boom, Alessandro Bertolini, Eric Hennes and Johannes F. J. van den Brand
Sensors 2021, 21(7), 2566; https://doi.org/10.3390/s21072566 - 6 Apr 2021
Cited by 3 | Viewed by 2956
Abstract
We present a novel analysis of gas damping in capacitive MEMS transducers that is based on a simple analytical model, assisted by Monte-Carlo simulations performed in Molflow+ to obtain an estimate for the geometry dependent gas diffusion time. This combination provides results with [...] Read more.
We present a novel analysis of gas damping in capacitive MEMS transducers that is based on a simple analytical model, assisted by Monte-Carlo simulations performed in Molflow+ to obtain an estimate for the geometry dependent gas diffusion time. This combination provides results with minimal computational expense and through freely available software, as well as insight into how the gas damping depends on the transducer geometry in the molecular flow regime. The results can be used to predict damping for arbitrary gas mixtures. The analysis was verified by experimental results for both air and helium atmospheres and matches these data to within 15% over a wide range of pressures. Full article
(This article belongs to the Section Electronic Sensors)
►▼ Show Figures

Figure 1

Figure 1
<p>(<b>a</b>) Photograph of the MEMS accelerometer for which the gas damping is modeled. The central 12.7 mg proof mass can move along the y-direction as indicated by the arrow. (<b>b</b>) An electron microscope zoom-in of part of the MEMS sensing capacitors indicated by the boxed area in (<b>a</b>) along with the nominal etched dimensions. The edge of the proof mass is visible at the top of the image and can move along the indicated y-direction. (<b>c</b>) Schematic cross-section of the accelerometer. Significant squeezed-film damping occurs both in the capacitor structures in the top silicon layer (A) and in the bottom silicon layer (B).</p> Full article ">Figure 2
<p>(<b>a</b>) Top view of typical 3D simulation result of a single set of capacitor plates (<math display="inline"><semantics> <mrow> <msub> <mi>d</mi> <mn>0</mn> </msub> <mo>=</mo> <mn>30</mn> <mo> </mo> <mrow> <mi mathvariant="sans-serif">μ</mi> <mi mathvariant="normal">m</mi> </mrow> </mrow> </semantics></math>). Particles spawn at one of the plates (blue dots) and are tracked through their reflections (green dots) until they leave the cavity (red dots). (<b>b</b>) Cross-section view SS’.</p> Full article ">Figure 3
<p>Schematic overview of the electronic system used to read the MEMS proof mass position. A transformer converts externally generated 100 <math display="inline"><semantics> <mi mathvariant="normal">kHz</mi> </semantics></math> sine with amplitude <math display="inline"><semantics> <msub> <mi>V</mi> <mi>d</mi> </msub> </semantics></math> to a balanced signal driving the MEMS capacitor bridge. The residual bridge current is integrated by a charge-amplifier, and its output voltage is buffered and sent to an external lock-in amplifier for synchronous demodulation. The circuit produces a low-frequency signal <math display="inline"><semantics> <msub> <mi>V</mi> <mrow> <mi>e</mi> <mi>r</mi> <mi>r</mi> </mrow> </msub> </semantics></math> proportional to <math display="inline"><semantics> <mrow> <mo>Δ</mo> <mi>C</mi> </mrow> </semantics></math>. The signals at different parts of the circuit are indicated.</p> Full article ">Figure 4
<p>Molflow+ simulation results for the diffusion time <math display="inline"><semantics> <mi>τ</mi> </semantics></math> as a function of plate separation <math display="inline"><semantics> <msub> <mi>d</mi> <mn>0</mn> </msub> </semantics></math> for air and helium particles at 300 <math display="inline"><semantics> <mi mathvariant="normal">K</mi> </semantics></math>, escaping from regions A and B as indicated in <a href="#sensors-21-02566-f001" class="html-fig">Figure 1</a>c. The dashed curve corresponds to particles escaping from region A, but with a 100 <math display="inline"><semantics> <mi mathvariant="sans-serif">μ</mi> </semantics></math><math display="inline"><semantics> <mi mathvariant="normal">m</mi> </semantics></math> thick silicon layer.</p> Full article ">Figure 5
<p><span class="html-italic">Q</span> factor measurements for a MEMS accelerometer with a natural frequency of <math display="inline"><semantics> <mrow> <mn>183.3</mn> <mi mathvariant="normal">Hz</mi> </mrow> </semantics></math> in air and helium atmospheres at different pressures. The dotted curves represent the expected <span class="html-italic">Q</span> from the total modeled gas damping <math display="inline"><semantics> <msub> <mi>γ</mi> <mrow> <mi>t</mi> <mi>o</mi> <mi>t</mi> </mrow> </msub> </semantics></math> as a function of pressure. The model is valid for <math display="inline"><semantics> <mrow> <mi>Kn</mi> <mo>&gt;</mo> <mn>10</mn> </mrow> </semantics></math>. The pressure was measured with two capacitive gauges, a CMR261 (<math display="inline"><semantics> <mrow> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>1</mn> </mrow> </msup> <mi mathvariant="normal">mbar</mi> </mrow> </semantics></math> to <math display="inline"><semantics> <mrow> <msup> <mn>10</mn> <mrow> <mn>3</mn> </mrow> </msup> <mi mathvariant="normal">mbar</mi> </mrow> </semantics></math>) and a CMR365 (<math display="inline"><semantics> <mrow> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mi mathvariant="normal">mbar</mi> </mrow> </semantics></math> to <math display="inline"><semantics> <mrow> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>1</mn> </mrow> </msup> <mi mathvariant="normal">mbar</mi> </mrow> </semantics></math>), and a cold cathode gauge, IKR251 (<math display="inline"><semantics> <mrow> <mo>&lt;</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mi mathvariant="normal">mbar</mi> </mrow> </semantics></math>).</p> Full article ">Figure 6
<p><span class="html-italic">Q</span> factors in air as a function of proof mass displacement for a MEMS accelerometer with <math display="inline"><semantics> <mrow> <msub> <mi>f</mi> <mn>0</mn> </msub> <mo>=</mo> <mn>164.0</mn> <mi>Hz</mi> </mrow> </semantics></math>.</p> Full article ">
19 pages, 2810 KiB  
Article
A Comparative Study of BDS Triple-Frequency Ambiguity Fixing Approaches for RTK Positioning
by Huizhong Zhu, Yangyang Lu, Longjiang Tang, Jun Li, Aigong Xu and Maorong Ge
Sensors 2021, 21(7), 2565; https://doi.org/10.3390/s21072565 - 6 Apr 2021
Cited by 4 | Viewed by 2570
Abstract
Concerning the triple-frequency ambiguity resolution, in principle there are three different realizations. The first one is to fix all the ambiguities of the original frequencies together. However, it is also believed that fixing the combined integer ambiguities with longer wavelength, such as extra-wide-lane [...] Read more.
Concerning the triple-frequency ambiguity resolution, in principle there are three different realizations. The first one is to fix all the ambiguities of the original frequencies together. However, it is also believed that fixing the combined integer ambiguities with longer wavelength, such as extra-wide-lane (EWL), wide-lane (WL), should be advantageous. Also, it is demonstrated that fixing sequentially EWL, WL and one type of original ambiguities provides better results, as the previously fixed ambiguities increase parameters’ precision for later fixings. In this paper, we undertake a comparative study of the three fixing approaches by means of experimental validation. In order to realize the three fixing approaches from the same information in terms of adjustment, we developed a processing strategy to provide fully consistent normal equations. We first generate the normal equation with the original undifferentiated carrier phase ambiguities, then map it into that with the combined and double-differenced ambiguities required by the individual approach for fixing. Four baselines of 258 m, 22 km, 47 km and 53 km are selected and processed in both static and kinematic mode using the three ambiguity-fixing approaches. Indicators including time of first fixed solution (TFFS), the correct fixing rate, positioning accuracy and RATIO are used to evaluate and investigate results. We also made a preliminary theoretical explanation of the results by looking into the decorrelation procedure of the ambiguity searching algorithm and the intermediate results. As conclusions, integrated searching of original ambiguities or combined ambiguities has almost the same fixing performance, whereas the sequential fixing of EWL, WL and B1 ambiguities overperforms the integrated searching. By the way, the third-frequency data can shorten the TFFS significantly but can hardly improve the positioning. Full article
(This article belongs to the Special Issue Signal Processing for GPS/GNSS/APNT Systems)
►▼ Show Figures

Figure 1

Figure 1
<p>RMS of the posterior phase residuals for each individual satellite and frequencies of the float solutions of the short Baseline A (<b>top</b>) and C1 (<b>bottom</b>) processed in static mode.</p> Full article ">Figure 2
<p>The top-left panel shows the estimated tropospheric zenith delay, and the other panels illustrated the ionospheric delays derived from B1/B2 phase observations aligned to the and the estimated slant ionospheric delay (red) for all the satellites over baseline C1 processed in kinematic mode in 10-min sessions. The estimates before ambiguity-fixing are not plotted.</p> Full article ">Figure 3
<p>Time series of ambiguity-fixing indicator RATIO of the fixing scheme S1 and S2 (<b>top</b>), and their differences (<b>middle</b>), and that of EWL and WL fixing of S3 (<b>bottom</b>), the RATIO of B1 fixing in S3 is also shown at the top panel for the BDS-TF solutions of Baseline A in static mode.</p> Full article ">Figure 4
<p>Time series of ambiguity-fixing indicator RATIO of the fixing scheme S1 and S2 (<b>top</b>), and their differences (<b>middle</b>), and that of EWL and WL fixing of S3 (<b>bottom</b>), the RATIO of B1 fixing in S3 is also shown at the top panel for the BDS-TF solutions of Baseline C1 in kinematic mode.</p> Full article ">Figure 5
<p>Example of the Eigenvalues of the covariance matrix after Z transformation of S1 and S2 for Baseline A.</p> Full article ">Figure 6
<p>Element values of the covariance matrix after Z-transformation for S1 and S2 and S3 from left to right, respectively. S3 has three blocks for EWL, WL and B1 sequentially.</p> Full article ">Figure 7
<p>Number of conditions for ambiguity resolution.</p> Full article ">
20 pages, 8614 KiB  
Article
Low Field Optimization of a Non-Contacting High-Sensitivity GMR-Based DC/AC Current Sensor
by Cristian Mușuroi, Mihai Oproiu, Marius Volmer, Jenica Neamtu, Marioara Avram and Elena Helerea
Sensors 2021, 21(7), 2564; https://doi.org/10.3390/s21072564 - 6 Apr 2021
Cited by 19 | Viewed by 3896
Abstract
Many applications require galvanic isolation between the circuit where the current is flowing and the measurement device. While for AC, the current transformer is the method of choice, in DC and, especially for low currents, other sensing methods must be used. This paper [...] Read more.
Many applications require galvanic isolation between the circuit where the current is flowing and the measurement device. While for AC, the current transformer is the method of choice, in DC and, especially for low currents, other sensing methods must be used. This paper aims to provide a practical method of improving the sensitivity and linearity of a giant magnetoresistance (GMR)-based current sensor by adapting a set of design rules and methods easy to be implemented. Our approach utilizes a multi-trace current trace and a double differential GMR based detection system. This essentially constitutes a planar coil which would effectively increase the usable magnetic field detected by the GMR sensor. An analytical model is developed for calculating the magnetic field generated by the current in the GMR sensing area which showed a significant increase in sensitivity up to 13 times compared with a single biased sensor. The experimental setup can measure both DC and AC currents between 2–300 mA, with a sensitivity between 15.62 to 23.19 mV/mA, for biasing fields between 4 to 8 Oe with a detection limit of 100 μA in DC and 100 to 300 μA in AC from 10 Hz to 50 kHz. Because of the double differential setup, the detection system has a high immunity to external magnetic fields and a temperature drift of the offset of about −2.59 × 10−4 A/°C. Finally, this setup was adapted for detection of magnetic nanoparticles (MNPs) which can be used to label biomolecules in lab-on-a-chip applications and preliminary results are reported. Full article
(This article belongs to the Special Issue Magnetic Sensors 2021)
►▼ Show Figures

Figure 1

Figure 1
<p>Working principle of the non-contacting current measurement setup utilizing current traces and a GMR based chip: (<b>a</b>) Multi-trace plane section; (<b>b</b>) Single trace plane section; (<b>c</b>) Multi-trace cross section; (<b>d</b>) Single trace cross section.</p> Full article ">Figure 2
<p>(<b>a</b>) Illustration of the geometry and parameters used in the analytical model to compute the magnetic field present in the sensor area. Note that the model takes into account that there is an odd number of traces that generate the magnetic field and the central trace is denoted as <span class="html-italic">n</span> = 0; (<b>b</b>) plane view of the planar coil with seven traces as designed for the actual implementation. <math display="inline"><semantics> <mover accent="true"> <mi>B</mi> <mo>→</mo> </mover> </semantics></math> is the resulting field in the central position (the middle trace is centered just below the sensitive area of the sensor).</p> Full article ">Figure 3
<p>Total magnetic field induction in the GMR sensor area (for I = 0.5 A) dependency on the number of current traces. The asymptotic tendency of Btot for the case where <span class="html-italic">D</span> = 0.22 mm is 6.485 Oe (Case I) while for <span class="html-italic">D</span> = 0.35 mm is 4.849 Oe (Case II). In both cases, <span class="html-italic">h</span> = 0.8 mm. Note that the field multiplying effect is clearly visible. Graphs were made using data obtained with the analytical method.</p> Full article ">Figure 4
<p>Micromagnetic simulation of a GMR element: (<b>a</b>) Schematic illustration of a GMR structure (two ferromagnetic layers FM1, FM2 and nonmagnetic layer, NM) with three distinct states depending on the parallel (P) or antiparallel (AP) alignment of layers magnetization; (<b>b</b>) Typical field dependence of the structure magnetization along the Ox axis, M, obtained by micromagnetic simulations and the calculated GMR effect when <span class="html-italic">H<sub>appl</sub></span> is directed over Ox axis.</p> Full article ">Figure 5
<p>Typical measured magnetic field dependency for the AA003-02 GMR sensor. In this case, the sensor was supplied with 4.096 V; this voltage is generated by a very stable source which is part from the EI 1040 instrumentation amplifier.</p> Full article ">Figure 6
<p>(<b>a</b>) GMR Testboard (custom PCB) optimized for low field detection (current measurement) using GMR sensors. The PCB is based on the 7 traces, Case II from <a href="#sensors-21-02564-f003" class="html-fig">Figure 3</a> and integrates the sensors, biasing coils, multi-turn planar coil, filtering system (capacitors for the INA 118 amplifiers power supply to filter any high frequency components and for the output of the amplifiers when measuring an AC signal to filter any DC component). (<b>b</b>) The functional block diagram of the current measurement system as well as the amplifier and data acquisition setup; notice the “U”-shaped structure of the circuit which produces the magnetic field which is applied to the sensors is integrated through the spiral trace which constitutes the planar coil.</p> Full article ">Figure 7
<p>Subblock components of the entire setup. The GMR sensors, temperature sensor, as well as the INA118 amplifiers are supplied from the EI1040 instrumentation amplifier, which also allows a variable gain to be set as needed. On the GMR Testboard, the differential output from each sensor is amplified by a fixed gain of 10. The output voltage from the LM15AZ temperature sensor is sent directly to the DAQ board. Also note that each coil is supplied separately from the HM8143 (parallel configuration), as this allow calibration of the biasing field for each GMR sensor.</p> Full article ">Figure 8
<p>Response of the system for a ±150 mA, DC current, <span class="html-italic">H</span><sub>bias</sub> was set to 8 Oe: (<b>a</b>) individual sensors response; (<b>b</b>) differential output.</p> Full article ">Figure 9
<p>DC response of the system for low current values: (<b>a</b>) Differential output for a ±5 mA, DC current, <span class="html-italic">H<sub>bias</sub></span> was set to 8 Oe; (<b>b</b>) Differential output of sensors polarized at 4,6,8 Oe, DC, ±2 mA (in this case, the adjusted R-squared for the fit function was: adjusted <span class="html-italic">R-square</span><sub>4Oe</sub> = 0.99961, adjusted <span class="html-italic">R-square</span><sub>6Oe</sub> = 0.9995, adjusted <span class="html-italic">R-square</span><sub>8Oe</sub> = 0.99943). Notice the linear characteristic of the output, although, for very low current values, some neliniarities can be present, but the overall linear tendency maintains.</p> Full article ">Figure 10
<p>AC response of the system, 8 Oe biased sensors, 1000 Hz sine waveform at 10 mA: (<b>a</b>) Differential output and trace current time dependency; (<b>b</b>) Differential output of sensors. The sensitivity for the sensors in the case is <span class="html-italic">S</span> = 0.2228 mV/mA ±9.6 ×10<sup>−5</sup>. Notice the sensitivity level is similar to that of DC measurements.</p> Full article ">Figure 11
<p>(<b>a</b>) AC response of the system, 8 Oe biased sensors, 1 kHz square waveform at 20 mA: a rise time and fall time of 15 µs was found in this case (measured between the 10–90% levels); The sensitivity for the sensors in this case is <span class="html-italic">S</span> = 0.2120 ± 7.186 × 10<sup>−4</sup> mV/mA; (<b>b</b>) AC response of the system, 8 Oe biased sensors, AC, 1 kHz, 20 mA pulse.</p> Full article ">Figure 12
<p>(<b>a</b>) AC frequency response (transfer function) of the system, 6 Oe biased sensors for different measured currents; (<b>b</b>) AC frequency response (transfer function) of the system, 8 Oe biased sensors for different measured currents.</p> Full article ">Figure 13
<p>(<b>a</b>) Impedance-frequency characteristic and current-voltage phase shift angle frequency dependency; (<b>b</b>) AC calibration curve in the 0–100 mA range for a 100 Hz sinewave. Note that the minimum trace current represented on the calibration curve is 1 mA.</p> Full article ">Figure 14
<p>(<b>a</b>) Measurement system adapted for the detection of MNP (<b>b</b>) Waveform of the applied biasing field, <span class="html-italic">H<sub>bias</sub></span> and output signals from setup: with standardized distilled water probe and MNP aqueous solution over the sensor’s surface after different elapsed times. The greatest field contribution of the MNPs was found at the <span class="html-italic">H<sub>bias</sub></span> = 4 Oe level where a ΔU = 0.0754 V signal variation was found compared with the case with no MNPs. The total gain of the signal was <span class="html-italic">G</span> = 100.</p> Full article ">
18 pages, 2012 KiB  
Article
Find Outliers of Image Edge Consistency by Weighted Local Linear Regression with Equality Constraints
by Mingzhu Zhu, Yaoqing Hu, Junzhi Yu, Bingwei He and Jiantao Liu
Sensors 2021, 21(7), 2563; https://doi.org/10.3390/s21072563 - 6 Apr 2021
Cited by 2 | Viewed by 2621
Abstract
In this paper, we propose a general method to detect outliers from contaminated estimates of various image estimation applications. The method does not require any prior knowledge about the purpose, theory or hardware of the application but simply relies on the law of [...] Read more.
In this paper, we propose a general method to detect outliers from contaminated estimates of various image estimation applications. The method does not require any prior knowledge about the purpose, theory or hardware of the application but simply relies on the law of edge consistency between sources and estimates. The method is termed as ALRe (anchored linear residual) because it is based on the residual of weighted local linear regression with an equality constraint exerted on the measured pixel. Given a pair of source and contaminated estimate, ALRe offers per-pixel outlier likelihoods, which can be used to compose the data weights of post-refinement algorithms, improving the quality of refined estimate. ALRe has the features of asymmetry, no false positive and linear complexity. Its effectiveness is verified on four applications, four post-refinement algorithms and three datasets. It demonstrates that, with the help of ALRe, refined estimates are better in the aspects of both quality and edge consistency. The results are even comparable to model-based and hardware-based methods. Accuracy comparison on synthetic images shows that ALRe could detect outliers reliably. It is as effective as the mainstream weighted median filter at spike detection and is significantly better at bad region detection. Full article
(This article belongs to the Section Sensing and Imaging)
►▼ Show Figures

Figure 1

Figure 1
<p>Bad pixel residuals of linear regression with and without equality constraint. Marked as ALRe and LRe, respectively. (<b>a</b>–<b>d</b>) Mean and minimal residuals of bad pixels with different biases; (<b>e</b>) Mean residuals of bad pixels under different configurations of bad pixel ratio; (<b>f</b>) Under different bad pixel ratios, the thresholds that pixels with smaller biases might have zero residuals.</p> Full article ">Figure 2
<p>Transmission map refinements. Transmission and weight maps are displayed in color. The warmer the color, the larger the value. (<b>a</b>,<b>b</b>) Hazy images and the initial transmission maps based on dark channel prior [<a href="#B27-sensors-21-02563" class="html-bibr">27</a>] and haze-line model [<a href="#B24-sensors-21-02563" class="html-bibr">24</a>] respectively; (<b>c</b>,<b>d</b>) The weight maps and the refined results based on Zhu et al. [<a href="#B28-sensors-21-02563" class="html-bibr">28</a>] and Berman et al. [<a href="#B24-sensors-21-02563" class="html-bibr">24</a>] respectively; (<b>e</b>) The weight maps based on ALRe and corresponding results.</p> Full article ">Figure 3
<p>Disparity map refinements. Disparity and weight maps are displayed in color. The warmer the color, the larger the value. (<b>a</b>) One of the two stereo images; (<b>b</b>) the weight map based on cross check; (<b>c</b>) the weight map based on ALRe (the segmentation threshold is 0.2); (<b>d</b>) initial disparity map based on Hosni et al. [<a href="#B25-sensors-21-02563" class="html-bibr">25</a>]; (<b>e</b>) the refined map based on (<b>b</b>); (<b>f</b>) the refined map based on (<b>c</b>).</p> Full article ">Figure 4
<p>Depth map refinements. Depth and weight maps are displayed in color. The warmer the color, the larger the value. (<b>a</b>) Color image; (<b>b</b>) rough depth map; (<b>c</b>) the weight map based on ALRe; (<b>d</b>) the refined result based on WMF without ALRe; (<b>e</b>) corresponding result with ALRe.</p> Full article ">Figure 5
<p>Feathering. Weight map is displayed in color. The warmer the color, the larger the value. (<b>a</b>) Input image; (<b>b</b>) rough mask; (<b>c</b>) the weight map based on ALRe; (<b>d</b>) the refined result based on GF without ALRe; (<b>e</b>) corresponding result with ALRe.</p> Full article ">Figure 6
<p>Edge preserving filtering. Weight map is displayed in color. The warmer the color, the larger the value. (<b>a</b>) Input image; (<b>b</b>) the smoothed input image based on Gaussian low-pass filter; (<b>c</b>) the weight map of (<b>b</b>) based on ALRe; (<b>d</b>) the smoothed result of (<b>a</b>) guided by itself based on GF; (<b>e</b>) the enhanced result of (<b>b</b>) guided by (<b>a</b>) based on GF with ALRe.</p> Full article ">Figure 7
<p>Effect of ALRe on WLS in transmission refinement. The ends of the orange line are the mean errors of rough and refined outputs of Zhu et al. [<a href="#B28-sensors-21-02563" class="html-bibr">28</a>]. The ends of the blue lines are the errors of WLS with and without ALRe.</p> Full article ">Figure 8
<p>Effect of ALRe on WMF in disparity post-refinement. The errors of cost-volume method [<a href="#B25-sensors-21-02563" class="html-bibr">25</a>] are marked by the dots with sample names. Each first linked dot is the error of WMF, and the next dot is the error of WMF with ALRe.</p> Full article ">Figure 9
<p>Disparity post-refinement using WMF with and without ALRe. Disparity and weight maps are displayed in color. The warmer the color, the larger the value. (<b>a</b>) One of the two stereo images; (<b>b</b>) the rough output of cost-volume [<a href="#B25-sensors-21-02563" class="html-bibr">25</a>]; (<b>c</b>) the refined result of (<b>b</b>) guided by (<b>a</b>) based on WMF without ALRe; (<b>d</b>) corresponding result with ALRe.</p> Full article ">Figure 10
<p>Effect of ALRe on matting. The errors of initial masks are marked by sample names. Each first linked dot is the error of GF, and the next is the error of GF with ALRe.</p> Full article ">Figure 11
<p>Feathering. (<b>a</b>) Color image; (<b>b</b>) the result of GF; (<b>c</b>) the result of GF with dilated foreground region; (<b>d</b>) the trimap; (<b>e</b>,<b>f</b>) corresponding results with ALRe.</p> Full article ">Figure 12
<p>Outlier detection accuracy. Left, spike detection; Middle, bad region detection; Right, the effects of source contrast and outlier strength.</p> Full article ">Figure 13
<p>Comparison of ALRe and WMF on bad region detection. (<b>a</b>) Source image; (<b>b</b>) contaminated estimate; (<b>c</b>) detection result of WMF, IoU = 0.6508; (<b>d</b>) detection result of ALRe, IoU = 0.9027; (<b>e</b>) ground truth.</p> Full article ">
21 pages, 55664 KiB  
Article
Application of the Gaussian Mixture Model to Classify Stages of Electrical Tree Growth in Epoxy Resin
by Abdullahi Abubakar Mas’ud, Arunachalam Sundaram, Jorge Alfredo Ardila-Rey, Roger Schurch, Firdaus Muhammad-Sukki and Nurul Aini Bani
Sensors 2021, 21(7), 2562; https://doi.org/10.3390/s21072562 - 6 Apr 2021
Cited by 5 | Viewed by 2970
Abstract
In high-voltage (HV) insulation, electrical trees are an important degradation phenomenon strongly linked to partial discharge (PD) activity. Their initiation and development have attracted the attention of the research community and better understanding and characterization of the phenomenon are needed. They are very [...] Read more.
In high-voltage (HV) insulation, electrical trees are an important degradation phenomenon strongly linked to partial discharge (PD) activity. Their initiation and development have attracted the attention of the research community and better understanding and characterization of the phenomenon are needed. They are very damaging and develop through the insulation material forming a discharge conduction path. Therefore, it is important to adequately measure and characterize tree growth before it can lead to complete failure of the system. In this paper, the Gaussian mixture model (GMM) has been applied to cluster and classify the different growth stages of electrical trees in epoxy resin insulation. First, tree growth experiments were conducted, and PD data captured from the initial to breakdown stage of the tree growth in epoxy resin insulation. Second, the GMM was applied to categorize the different electrical tree stages into clusters. The results show that PD dynamics vary with different stress voltages and tree growth stages. The electrical tree patterns with shorter breakdown times had identical clusters throughout the degradation stages. The breakdown time can be a key factor in determining the degradation levels of PD patterns emanating from trees in epoxy resin. This is important in order to determine the severity of electrical treeing degradation, and, therefore, to perform efficient asset management. The novelty of the work presented in this paper is that for the first time the GMM has been applied for electrical tree growth classification and the optimal values for the hyperparameters, i.e., the number of clusters and the appropriate covariance structure, have been determined for the different electrical tree clusters. Full article
(This article belongs to the Section Sensor Materials)
►▼ Show Figures

Figure 1

Figure 1
<p>Test circuit for electrical tree growth experiments.</p> Full article ">Figure 2
<p>Time series of partial discharge (PD) amplitude and tree length for each sample: (<b>a</b>) Sample A, (<b>b</b>) Sample B, (<b>c</b>) Sample C, (<b>d</b>) Sample D, and (<b>e</b>) Sample E different.</p> Full article ">Figure 3
<p>Picture of Sample B at interval 2, showing a filamentary tree growth.</p> Full article ">Figure 4
<p>Images of electrical trees at interval 6 in each sample: (<b>a</b>) Sample A, (<b>b</b>) Sample B, (<b>c</b>) Sample C, (<b>d</b>) Sample D, and (<b>e</b>) Sample E.</p> Full article ">Figure 5
<p>Procedure to choose the hyperparameters.</p> Full article ">Figure 6
<p>Bar Plot of Akaike information criterion values for each fit in dataset A1.</p> Full article ">Figure 7
<p>Bar Plot of Bayesian information criterion values for each fit in dataset A1.</p> Full article ">Figure 8
<p>Plot of Akaike information criterion values for each fit in dataset A1.</p> Full article ">Figure 9
<p>Plot of Bayesian information criterion values for each fit in dataset A1.</p> Full article ">Figure 10
<p>Clusters for the different datasets at the initial stage of degradation (<b>a</b>) Sample A, (<b>b</b>) Sample B, (<b>c</b>) Sample C, (<b>d</b>) Sample D, and (<b>e</b>) Sample E.</p> Full article ">Figure 11
<p>Clusters for the different samples at the final stage (interval 10): (<b>a</b>) Sample A (<b>b</b>) Sample B (<b>c</b>) Sample C (<b>d</b>) Sample D and (<b>e</b>) Sample E.</p> Full article ">Figure 12
<p>Flowchart for GMM Electrical tree pattern recognition.</p> Full article ">
27 pages, 6897 KiB  
Article
Indirect Temperature Measurement in High Frequency Heating Systems
by Alexander Oskolkov, Igor Bezukladnikov and Dmitriy Trushnikov
Sensors 2021, 21(7), 2561; https://doi.org/10.3390/s21072561 - 6 Apr 2021
Cited by 7 | Viewed by 4977
Abstract
One of the biggest challenges of fused deposition modeling (FDM)/fused filament fabrication (FFF) 3D-printing is maintaining consistent quality of layer-to-layer adhesion, and on the larger scale, homogeneity of material inside the whole printed object. An approach for mitigating and/or resolving those problems, based [...] Read more.
One of the biggest challenges of fused deposition modeling (FDM)/fused filament fabrication (FFF) 3D-printing is maintaining consistent quality of layer-to-layer adhesion, and on the larger scale, homogeneity of material inside the whole printed object. An approach for mitigating and/or resolving those problems, based on the rapid and reliable control of the extruded material temperature during the printing process, was proposed. High frequency induction heating of the nozzle with a minimum mass (<1 g) was used. To ensure the required dynamic characteristics of heating and cooling processes in a high power (peak power > 300 W) heating system, an indirect (eddy current) temperature measurement method was proposed. It is based on dynamic analysis over various temperature-dependent parameters directly in the process of heating. To ensure better temperature measurement accuracy, a series-parallel resonant circuit containing an induction heating coil, an approach of desired signal detection, algorithms for digital signal processing and a regression model that determines the dependence of the desired signal on temperature and magnetic field strength were proposed. The testbed system designed to confirm the results of the conducted research showed the effectiveness of the proposed indirect measurement method. With an accuracy of ±3 °C, the measurement time is 20 ms in the operating temperature range from 50 to 350 °C. The designed temperature control system based on an indirect measurement method will provide high mechanical properties and consistent quality of printed objects. Full article
(This article belongs to the Special Issue Next-Generation Temperature Sensors)
►▼ Show Figures

Figure 1

Figure 1
<p>(<b>a</b>) Appearance of the ferromagnetic nozzle, (<b>b</b>) appearance of the pilot extruder.</p> Full article ">Figure 2
<p>Functional block diagram of the testbed system: (<b>1</b>) the induction heating coil; (<b>2</b>) the nozzle; (<b>3</b>) high frequency (HF) inverter; (<b>4</b>) the sensing coil; (<b>5</b>) the unit for recording and processing a measuring signal (ARM-microcontroller).</p> Full article ">Figure 3
<p>Appearance of the testbed system.</p> Full article ">Figure 4
<p>Multisim model of LCL series-parallel resonant circuit.</p> Full article ">Figure 5
<p>Frequency range and corresponding phase shift between inductor (L1) current and output voltage. Nozzle temperature (T) for resonance condition. (<b>a</b>) T = 200 °C, (<b>b</b>) T = 25 °C, (<b>c</b>) T = 400 °C.</p> Full article ">Figure 5 Cont.
<p>Frequency range and corresponding phase shift between inductor (L1) current and output voltage. Nozzle temperature (T) for resonance condition. (<b>a</b>) T = 200 °C, (<b>b</b>) T = 25 °C, (<b>c</b>) T = 400 °C.</p> Full article ">Figure 6
<p>High-frequency induction heating system with LCL-resonant output and Sensing circuit.</p> Full article ">Figure 7
<p>Desired signal amplitude (error) as a function of time (ticks): (<b>a</b>) power consumption of 80%, (<b>b</b>) power consumption of 90%.</p> Full article ">Figure 8
<p>Desired signal amplitude (error) as a function of time (ticks) after filtration: (<b>a</b>) power consumption of 80%, (<b>b</b>) power consumption of90%.</p> Full article ">Figure 9
<p>Ideal heating curve of the nozzle with power consumption of 90%.</p> Full article ">Figure 10
<p>Estimated regression equation and observed values of “error”.</p> Full article ">Figure 11
<p>Desired signal amplitude (error) dependency on magnetic field strength (phase shift if<sub>i</sub>) with fixed nozzle temperature: (<b>a</b>) 110 °C, (<b>b</b>) 750 °C.</p> Full article ">Figure 11 Cont.
<p>Desired signal amplitude (error) dependency on magnetic field strength (phase shift if<sub>i</sub>) with fixed nozzle temperature: (<b>a</b>) 110 °C, (<b>b</b>) 750 °C.</p> Full article ">Figure 12
<p>Observed versus predicted values of “error”. Dependence of “error” on temperature with fixed power consumption: (<b>a</b>) 0.2%, (<b>b</b>) 20%.</p> Full article ">Figure 12 Cont.
<p>Observed versus predicted values of “error”. Dependence of “error” on temperature with fixed power consumption: (<b>a</b>) 0.2%, (<b>b</b>) 20%.</p> Full article ">Figure 13
<p>(<b>a</b>) Appearance of the test product surface printed of polyamide (PA); (<b>b</b>) appearance of the showpiece printed from polyetheretherketone (PEEK).</p> Full article ">Figure 14
<p>(<b>a</b>) Appearance of the fractured tensile specimen printed of acrylonitrile butadiene styrene (ABS); (<b>b</b>) appearance of the fractured tensile specimen printed of polylactic acid (PLA).</p> Full article ">Figure 15
<p>Optical microscopy of the cross-sectional surfaces of the FDM-printed specimens after tensile testing:(<b>a</b>) fractured tensile specimens printed of ABS; (<b>b</b>)fractured tensile specimens printed of PLA.</p> Full article ">Figure 16
<p>Cooling time curves of the induction heated (proposed)nozzle at various extrusion speeds and of the conventional hotend assembly (nozzle+heating block).</p> Full article ">
15 pages, 4050 KiB  
Article
Construction of Ultrasonic Tactile Force Feedback Model in Teleoperation Robot System
by Yang Liu, Xiaoling Li, Jiarui Lai, Ziming Zheng, Huijin Zhu and Min Li
Sensors 2021, 21(7), 2560; https://doi.org/10.3390/s21072560 - 6 Apr 2021
Cited by 10 | Viewed by 3506
Abstract
The ultrasonic phased array as an emerging interactive tool is increasingly used for aerial tactile interaction. However, there is almost no method to achieve remote variable force feedback through the ultrasonic phased array as far as we know. This article presents a force [...] Read more.
The ultrasonic phased array as an emerging interactive tool is increasingly used for aerial tactile interaction. However, there is almost no method to achieve remote variable force feedback through the ultrasonic phased array as far as we know. This article presents a force tactile feedback method for teleoperating robot systems that tracks the five fingers and forms a focus on the fingertips. First, the perceived size of the focus depends on the input parameters. The influence of the parameters on the physical output pressure intensity was obtained through physical test experiments. Then, the absolute threshold and difference threshold of human perception were studied through psychophysical experimental methods. Finally, the input parameters were selected according to the experimental results. According to the collected data, the construction of the force regression model was completed, and different parameters were mapped to the perceived intensity. The contact force generated in the actual operation is fed back to the haptic system, and the constructed model automatically adjusts the control parameters to ensure that the user’s hand presents a sensory output corresponding to the intensity change. The entire force feedback system is evaluated, and results show that the system shows good perceptual quality. Full article
(This article belongs to the Section Physical Sensors)
►▼ Show Figures

Figure 1

Figure 1
<p>Conceptual diagram of teleoperation robot system.</p> Full article ">Figure 2
<p>The flow chart of tactile points focusing.</p> Full article ">Figure 3
<p>The overall flowchart of our proposed system. It includes the human-computer interaction terminal and the control terminal, as well as the method of data transmission and the method of realizing device control.</p> Full article ">Figure 4
<p>Measurement setup: XYZ workbench, data collector, acoustic sensor, ultrasonic mid-air haptic display.</p> Full article ">Figure 5
<p>The relationship between input command intensity and output intensity.</p> Full article ">Figure 6
<p>The relationship between the distance from centre point and the output intensity.</p> Full article ">Figure 7
<p>The relationship between height and output intensity.</p> Full article ">Figure 8
<p>The result of the absolute threshold measurement. An example: absolute threshold measurement result of two staircase programs.</p> Full article ">Figure 9
<p>Psychophysical adaptive staircase for subject 2 with a reference flow rate of 0.5.</p> Full article ">Figure 10
<p>Three-dimensional display of test data points. The <span class="html-italic">x</span>-axis is the intensity of the input command, the <span class="html-italic">y</span>-axis is the height, the <span class="html-italic">z</span>-axis is the distance from the centre point, the test points are characteristic data points of different parameter combinations, and the measured pressure is indicated by color.</p> Full article ">Figure 11
<p>Frame diagram of real-time force feedback system.</p> Full article ">Figure 12
<p>Test and verification experiment diagram in teleoperation system.</p> Full article ">Figure 13
<p>The predicted and measured values at different positions from the centre point on the horizontal plane. The value indicated by the arrow is the relative error.</p> Full article ">Figure 14
<p>Test values and predicted values at different heights on the vertical line where the centre point of the array is located.</p> Full article ">
16 pages, 2764 KiB  
Article
Potentiometric Performance of a Highly Flexible-Shaped Trifunctional Sensor Based on ZnO/V2O5 Microrods
by Alfred Bekoe Appiagyei and Jeong In Han
Sensors 2021, 21(7), 2559; https://doi.org/10.3390/s21072559 - 6 Apr 2021
Cited by 5 | Viewed by 2630
Abstract
A trifunctional flexible sensor was fabricated on a polyethylene terephthalate (PET) fiber surface. Synthesized ZnO and ZnO/V2O5 composite were coated on ZnO seed layer sputtered PET fiber. X-ray diffraction (XRD) and photoelectron spectroscopy (XPS) techniques confirmed the exact formation of [...] Read more.
A trifunctional flexible sensor was fabricated on a polyethylene terephthalate (PET) fiber surface. Synthesized ZnO and ZnO/V2O5 composite were coated on ZnO seed layer sputtered PET fiber. X-ray diffraction (XRD) and photoelectron spectroscopy (XPS) techniques confirmed the exact formation of ZnO and ZnO/V2O5. The fabricated ZnO/V2O5 on ZnO seeds base temperature sensor recorded better electrical properties and reversibility with a maximum temperature coefficient resistance (TCR) of 0.0111 °C−1. A calibration curve (R = 0.9941) within glucose concentration of (10 µM–10 mM) was obtained at +0.8 V vs. Ag/AgCl from current-voltage curves which assisted in calculating glucose sensitivity, limit of detection (LOD), limit of quantification (LOQ). The electrode achieved an outstanding performance of sensitivity (72.06 µAmM−1cm−2), LOD (174 µM), and LOQ (582 µM) at optimum deposition time. Interference from oxidation of interfering biomolecules such as ascorbic acid, dopamine, and uric acid were negligible compared to glucose. Finally, the fabricated electrode was employed as a pH sensor and displayed a pH sensitivity of 42.26 mV/pH (R = 0.9922). This fabricated ZnO/V2O5 electrode exhibited high sensitivity and a stable combined temperature, glucose, and pH sensor which is promising for development of multifunctional sensors in next generation wearables. Full article
(This article belongs to the Section Sensor Materials)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Photographic image of ZnO/V<sub>2</sub>O<sub>5</sub> deposited PET single fiber (<b>a</b>) at straight and (<b>b</b>) fullbending conditions.</p> Full article ">Figure 2
<p>FESEM images of ZnO seed layer (<b>a1</b>,<b>b1</b>,<b>c1</b>); ZnO microrods (<b>a2</b>,<b>b2</b>,<b>c2</b>) and ZnO/V<sub>2</sub>O<sub>5</sub> composite (<b>a3</b>,<b>b3</b>,<b>c3</b>) for deposition time of 10 min, 20 min, and 30 min respectively.</p> Full article ">Figure 3
<p>XRD patterns of ZnO/V<sub>2</sub>O<sub>5</sub> with seed layer deposition time of (<b>a</b>) 10 min, (<b>b</b>) 20 min, and (<b>c</b>) 30 min.</p> Full article ">Figure 4
<p>XPS spectra of ZnO/V<sub>2</sub>O<sub>5</sub> with seed layer deposition time of (<b>a</b>) 10 min, (<b>b</b>) 20 min, and (<b>c</b>) 30 min.</p> Full article ">Figure 5
<p>(<b>a</b>) Variation of the electrical properties of ZnO/V<sub>2</sub>O<sub>5</sub> temperature sensor at different temperature conditions. Resistance–temperature relation of ZnO/V<sub>2</sub>O<sub>5</sub> temperature sensor for seed layer deposition time of (<b>b</b>) 10 min, (<b>c</b>) 20 min, and (<b>d</b>) 30 min.</p> Full article ">Figure 6
<p>(<b>a</b>) Cyclic voltammetry curve of ZnO/V<sub>2</sub>O<sub>5</sub> electrode at different scan rate applied in glucose sensing, (<b>b</b>) calibration plot of blank, ZnO and ZnO/V<sub>2</sub>O<sub>5</sub> glucose sensors at +0.8 V with a straight line representing the linear fit for ZnO seed layer deposited on PET for 20 min, (<b>c</b>) current-time response monitoring according to increasing glucose concentration towards ZnO/V<sub>2</sub>O<sub>5</sub> electrode, and (<b>d</b>) bar diagram representation of the interference effect at +1.0 V.</p> Full article ">Figure 7
<p>(<b>a</b>) Cyclic voltammetry curve of ZnO/V<sub>2</sub>O<sub>5</sub> electrode at different scan rates in NaOH/HCl mixture of pH 4, (<b>b</b>) calibration plot for ZnO and ZnO/V<sub>2</sub>O<sub>5</sub> pH sensors with a straight line representing the linear fit, (<b>c</b>) potential–time response obtained on increasing the pH of NaOH/HCl electrolyte at ZnO/V<sub>2</sub>O<sub>5</sub> electrode, and (<b>d</b>) calculated sensitivity after various bending cycles to validate the mechanical stability of ZnO/V<sub>2</sub>O<sub>5</sub> pH sensors.</p> Full article ">Scheme 1
<p>Representation of the fabrication of ZnO microrods and ZnO/V<sub>2</sub>O<sub>5</sub> composite on a PET fiber. (<b>a</b>) bare PET monofilament substrate (<b>b</b>) RF sputtered ZnO seeds on PET fiber. Hydrothermal synthesis of (<b>c</b>) ZnO microrods and (<b>d</b>) ZnO/V<sub>2</sub>O<sub>5</sub> composite on PET fiber.</p> Full article ">
13 pages, 1081 KiB  
Article
Oxygen Sensing of Pt/PEO-TiO2 in Humid Atmospheres at Moderate Temperatures
by Bernd Engelkamp and Klaus Schierbaum
Sensors 2021, 21(7), 2558; https://doi.org/10.3390/s21072558 - 6 Apr 2021
Cited by 3 | Viewed by 2352
Abstract
Here, we show that the presence of adsorbed water improves the oxygen-sensing properties of Pt/TiO2 at moderate temperatures. The studied interface is based on porous plasma electrolytic oxidized titanium (PEO-TiO2) covered with platinum clusters. The electrical resistance across Pt/PEO-TiO2 [...] Read more.
Here, we show that the presence of adsorbed water improves the oxygen-sensing properties of Pt/TiO2 at moderate temperatures. The studied interface is based on porous plasma electrolytic oxidized titanium (PEO-TiO2) covered with platinum clusters. The electrical resistance across Pt/PEO-TiO2 is explained by an electronic depletion layer. Oxygen adsorbates further increase the depletion by inducing extrinsic interface states, which are occupied by TiO2 conduction band electrons. The high oxygen partial pressure in ambient air substantially limits the electron transport across the interface. Our DC measurements at defined levels of humidity at 30 ∘C show that adsorbed water counteracts this shortcoming, allowing oxygen sensing at room conditions. In addition, response and recovery times from temporal oxygen exposure decrease with humidity. We attribute the effects to competing adsorption processes and reactions of water with adsorbed oxygen species and/or lattice oxygen, which involve electron re-injection to the TiO2 conduction band. Elevated temperatures up to 170 ∘C attenuate the effects, presumably due to the lower binding strength to the surface of molecular water compared with oxygen adsorbates. Full article
(This article belongs to the Section Sensor Materials)
►▼ Show Figures

Figure 1

Figure 1
<p>(<b>a</b>) SEM image of the Pt/PEO-TiO<math display="inline"><semantics> <msub> <mrow/> <mn>2</mn> </msub> </semantics></math> surface. (<b>b</b>) XPS spectra of the O 1s binding energy range for the PEO-TiO<math display="inline"><semantics> <msub> <mrow/> <mn>2</mn> </msub> </semantics></math> sample at room temperature. The first peak is centered around 530.1 eV and is assigned to O<math display="inline"><semantics> <msup> <mrow/> <mrow> <mn>2</mn> <mo>−</mo> </mrow> </msup> </semantics></math>, while the second peak is centered around 531.7 eV and is assumed to originate from OH groups.</p> Full article ">Figure 2
<p>Water adsorption/desorption isotherm recorded at 30 <math display="inline"><semantics> <mrow> <msup> <mrow/> <mo>∘</mo> </msup> <mi mathvariant="normal">C</mi> </mrow> </semantics></math> reveals insights into the water adsorption process on PEO-TiO<math display="inline"><semantics> <msub> <mrow/> <mn>2</mn> </msub> </semantics></math>. The adsorption isotherm is additionally fitted by the BET equation.</p> Full article ">Figure 3
<p>(<b>a</b>) The sample setup under the forward bias condition. Schematically shown are the relevant junctions, the electronic transport, and the possible contribution of ionic transport. (<b>b</b>) <span class="html-italic">I–V</span> curves were recorded at 30 <math display="inline"><semantics> <mrow> <msup> <mrow/> <mo>∘</mo> </msup> <mi mathvariant="normal">C</mi> </mrow> </semantics></math> and constant levels of humidity. The inner graph shows the ideality factor <span class="html-italic">n</span>, which was obtained by fitting the <span class="html-italic">I–V</span> curves according to the concept of an interface-controlled Schottky contact. The error bars are deduced from the fitting standard error.</p> Full article ">Figure 4
<p>Illustration of the band structure of Pt/TiO<math display="inline"><semantics> <msub> <mrow/> <mn>2</mn> </msub> </semantics></math> at three phase boundaries (TPBs). In the left scheme, the depletion and band bending is strongly promoted by the high surface density of ionosorbed oxygen, e.g., <math display="inline"><semantics> <msub> <mrow> <msup> <mi mathvariant="normal">O</mi> <mo>−</mo> </msup> </mrow> <mi>ads</mi> </msub> </semantics></math>. When the humidity increases, the density decreases. The impact on the band structure is shown in the right scheme. In comparison, the band bending is reduced by <math display="inline"><semantics> <mrow> <mi>e</mi> <mo>Δ</mo> <msub> <mi>V</mi> <mi mathvariant="normal">S</mi> </msub> </mrow> </semantics></math>.</p> Full article ">Figure 5
<p>(<b>a</b>) Time dependent current at 3 <math display="inline"><semantics> <mi mathvariant="normal">V</mi> </semantics></math> in constant levels of r.h. at 30 <math display="inline"><semantics> <mrow> <msup> <mrow/> <mo>∘</mo> </msup> <mi mathvariant="normal">C</mi> </mrow> </semantics></math> with exposure to different concentrations of oxygen. (<b>b</b>) The response time was determined by the time until the signal has fallen to 10% of the initial value relative to the minimum after five minutes of oxygen exposure. In analogous way, the subsequent recovery time was determined by the time until the signal has reached 90% of the following maximum (in nitrogen).</p> Full article ">Figure 6
<p>(<b>a</b>) Current transients at 3 <math display="inline"><semantics> <mi mathvariant="normal">V</mi> </semantics></math> in 90% r.h. at temperatures between 30 and 170 <math display="inline"><semantics> <mrow> <msup> <mrow/> <mo>∘</mo> </msup> <mi mathvariant="normal">C</mi> </mrow> </semantics></math>. The time with oxygen addition was 10 min and the time between oxygen exposure was 30 min. (<b>b</b>) The corresponding <span class="html-italic">I–V</span> curves were recorded immediately after the transients. The inner graph shows the dependence of the corresponding ideality factor <span class="html-italic">n</span> on temperature. The error bars are estimated based on fitting the standard error.</p> Full article ">
25 pages, 6804 KiB  
Article
Design of Minimum Nonlinear Distortion Reconfigurable Antennas for Next-Generation Communication Systems
by Germán Augusto Ramírez Arroyave, Antoni Barlabé, Lluís Pradell, Javier Leonardo Araque Quijano, Bedri A. Cetiner and Luis Jofre-Roca
Sensors 2021, 21(7), 2557; https://doi.org/10.3390/s21072557 - 6 Apr 2021
Cited by 8 | Viewed by 3399
Abstract
Nonlinear effects in the radio front-end can degrade communication quality and system performance. In this paper we present a new design technique for reconfigurable antennas that minimizes the nonlinear distortion and maximizes power efficiency through the minimization of the coupling between the internal [...] Read more.
Nonlinear effects in the radio front-end can degrade communication quality and system performance. In this paper we present a new design technique for reconfigurable antennas that minimizes the nonlinear distortion and maximizes power efficiency through the minimization of the coupling between the internal switching ports and the external feeding ports. As a nonlinear design and validation instance, we present the nonlinear characterization up to 50 GHz of a PIN diode commonly used as a switch for reconfigurable devices in the microwave band. Nonlinear models are extracted through X-parameter measurements supported by accurate calibration and de-embedding procedures. Nonlinear switch models are validated by S-parameter measurements in the low power signal regime and by harmonic measurements in the large-signal regime and are further used to predict the measured nonlinearities of a reconfigurable antenna. These models have the desired particularity of being integrated straightforwardly in the internal multi-port method formulation, which is used and extended to account for the power induced on the switching elements. A new figure of merit for the design of reconfigurable antennas is introduced—the power margin, that is, the power difference between the fed port and the switching elements, which combined with the nonlinear load models directly translates into nonlinearities and power-efficiency-related metrics. Therefore, beyond traditional antenna aspects such as port match, gain, and beam orientation, switch power criteria are included in the design methodology. Guidelines for the design of reconfigurable antennas and parasitic layers of minimum nonlinearity are provided as well as the inherent trade-offs. A particular antenna design suitable for 5G communications in the 3.5 GHz band is presented according to these guidelines, in which the specific switching states for a set of target performance metrics are obtained via a balancing of the available figures of merit with multi-objective separation criteria, which enables good control of the various design trade-offs. Average Error Vector Magnitude (EVM) and power efficiency improvement of 12 and 6 dB, respectively, are obtained with the application of this design approach. In summary, this paper introduces a new framework for the nonlinear modeling and design of reconfigurable antennas and provides a set of general-purpose tools applicable in cases beyond those used as examples and validation in this work. Additionally, the use of these models and guidelines is presented, demonstrating one of the most appealing advantages of the reconfigurable parasitic layer approach, their low nonlinearity. Full article
(This article belongs to the Special Issue Antenna Design for 5G and Beyond)
►▼ Show Figures

Figure 1

Figure 1
<p>Exploded view of the Parasitic-Layer-based Reconfigurable-antenna used as case study. (Units in mm).</p> Full article ">Figure 2
<p>X-parameters measurement setup.</p> Full article ">Figure 3
<p>Measured S-parameter (de-embedded) vs. Bias.</p> Full article ">Figure 4
<p>X-Parameters variation with input power at fundamental <span class="html-italic">f</span><sub>1</sub> = 5 GHz. (<span style="color:red">⚊</span> <span class="html-italic">f</span><sub>1</sub>, <span style="color:blue">⚊</span> 2<span class="html-italic">f</span><sub>1</sub>, <span style="color:#FF00FF">⚊</span> 3<span class="html-italic">f</span><sub>1</sub>, ⎯ “Off”, --- “On”).</p> Full article ">Figure 5
<p>S- and X- parameters small signal convergence.</p> Full article ">Figure 6
<p>Parasitic-Layer-based Reconfigurable-antenna and antenna equivalent circuit.</p> Full article ">Figure 7
<p>Port match in a test reconfigurable antenna.</p> Full article ">Figure 8
<p>Harmonic components generated by the test reconfigurable antenna, with indication of <math display="inline"><semantics> <mrow> <mi>I</mi> <msub> <mi>P</mi> <mn>3</mn> </msub> </mrow> </semantics></math>. (<b>left</b>) <math display="inline"><semantics> <msub> <mi>M</mi> <mn>1</mn> </msub> </semantics></math>, (<b>right</b>) <math display="inline"><semantics> <msub> <mi>M</mi> <mn>4</mn> </msub> </semantics></math>.</p> Full article ">Figure 9
<p>Error Vector Magnitude (EVM) of test reconfigurable antenna.</p> Full article ">Figure 10
<p>Flow chart with the main elements of the design methodology.</p> Full article ">Figure 11
<p>Parasitic-Layer-based Reconfigurable-antenna.</p> Full article ">Figure 12
<p>Parasitic-Layer-based Reconfigurable-antenna tradeoffs.</p> Full article ">Figure 13
<p>Design Trade-Offs of a Parasitic Layer (PL) based Reconfigurable Antenna (RA).</p> Full article ">Figure 14
<p>Optimization process.</p> Full article ">Figure 15
<p>Radiation diagrams at Port 1, for operating mode <span class="html-italic">M<sub>i</sub></span>, (<span class="html-italic">i</span> ∈ 1 ... 9), pointing towards (<span class="html-italic">θ</span><sub>0</sub>, <span class="html-italic">ϕ</span><sub>0</sub>).</p> Full article ">Figure 16
<p>EVM of the Minimum Nonlinearity Parasitic-Layer-based Reconfigurable Antenna.</p> Full article ">
20 pages, 1698 KiB  
Article
Qualitative Assessment of Effective Gamification Design Processes Using Motivators to Identify Game Mechanics
by Eva Villegas, David Fonseca, Enric Peña, Paula Bonet and Sara Fernández-Guinea
Sensors 2021, 21(7), 2556; https://doi.org/10.3390/s21072556 - 6 Apr 2021
Cited by 13 | Viewed by 5107
Abstract
This research focuses on the study and qualitative assessment of the relationships between motivators and game mechanics per the ratings of expert gamification consultants. By taking this approach, it is intended that during the design phase of a gamified system, decisions can be [...] Read more.
This research focuses on the study and qualitative assessment of the relationships between motivators and game mechanics per the ratings of expert gamification consultants. By taking this approach, it is intended that during the design phase of a gamified system, decisions can be made about the design of the system based on the motivators of each of the profiles. These motivators can be determined from the information provided by the potential players themselves. The research presented starts from a previous analysis in which, based on the three most used gamification frameworks and through a card sorting technique that allows the user to organize and classify the content, a set of mechanics are determined. In the present study, each of the mechanics is analyzed, and a more precise motive is decided. As a result, a higher level of personalization is achieved and, consequently, approximates a higher level of gamification effectiveness. The main conclusions are implemented in the development of the Game4City 3.0 project, which addresses gamified and interactive strategies to visualize urban environments in 3D at an educational and social level. Full article
(This article belongs to the Special Issue Pervasive Mobile-Based Games, AR/VR and Sensors)
►▼ Show Figures

Figure 1

Figure 1
<p>Representation of Core engagement loop model.</p> Full article ">Figure 2
<p>Previous evaluation with the Open Card Sorting Method: (<b>a</b>) image of the classification of mechanics made by User 2; (<b>b</b>) image of the classification of mechanics made by User 3.</p> Full article ">Figure 3
<p>Percentage of motives related to mechanics.</p> Full article ">
19 pages, 2872 KiB  
Article
Label-Free Protein Detection by Micro-Acoustic Biosensor Coupled with Electrical Field Sorting. Theoretical Study in Urine Models
by Nikolay Mukhin, Georgii Konoplev, Aleksandr Oseev, Marc-Peter Schmidt, Oksana Stepanova, Andrey Kozyrev, Alexander Dmitriev and Soeren Hirsch
Sensors 2021, 21(7), 2555; https://doi.org/10.3390/s21072555 - 6 Apr 2021
Cited by 6 | Viewed by 4259
Abstract
Diagnostic devices for point-of-care (POC) urine analysis (urinalysis) based on microfluidic technology have been actively developing for several decades as an alternative to laboratory based biochemical assays. Urine proteins (albumin, immunoglobulins, uromodulin, haemoglobin etc.) are important biomarkers of various pathological conditions and should [...] Read more.
Diagnostic devices for point-of-care (POC) urine analysis (urinalysis) based on microfluidic technology have been actively developing for several decades as an alternative to laboratory based biochemical assays. Urine proteins (albumin, immunoglobulins, uromodulin, haemoglobin etc.) are important biomarkers of various pathological conditions and should be selectively detected by urinalysis sensors. The challenge is a determination of different oligomeric forms of the same protein, e.g., uromodulin, which have similar bio-chemical affinity but different physical properties. For the selective detection of different types of proteins, we propose to use a shear bulk acoustic resonator sensor with an additional electrode on the upper part of the bioliquid-filled channel for protein electric field manipulation. It causes modulation of the protein concentration over time in the near-surface region of the acoustic sensor, that allows to distinguish proteins based on their differences in diffusion coefficients (or sizes) and zeta-potentials. Moreover, in order to improve the sensitivity to density, we propose to use structured sensor interface. A numerical study of this approach for the detection of proteins was carried out using the example of albumin, immunoglobulin, and oligomeric forms of uromodulin in model urine solutions. In this contribution we prove the proposed concept with numerical studies for the detection of albumin, immunoglobulin, and oligomeric forms of uromodulin in urine models. Full article
(This article belongs to the Collection Enabling Technologies for Biosensors)
►▼ Show Figures

Figure 1

Figure 1
<p>Top (<b>a</b>) and slice (<b>b</b>) view of a schematic model of the device for manipulating and detecting proteins: <span class="html-italic">1</span>—Biological liquid (urine) in the canal; <span class="html-italic">2</span>—filter; <span class="html-italic">3</span>—temperature control; <span class="html-italic">4</span>—<span class="html-italic">pH</span> control; <span class="html-italic">5</span>—piezoelectric quartz crystal resonators; <span class="html-italic">6</span>—bottom electrodes under alternating probing potential <span class="html-italic">U</span><sub>1</sub> and <span class="html-italic">U</span><sub>2</sub>; 7—top grounded electrode; <span class="html-italic">8</span>—set of protrusions placed on the top electrode surface; <span class="html-italic">9</span>—direction of shear vibrations of a quartz resonator, perpendicular to the microfluidic channels and a set of top electrode protrusions; <span class="html-italic">10</span>—addition electrode over the channel for particles manipulations in liquid due to the application of electric potential <span class="html-italic">U</span><sub>0</sub>; <span class="html-italic">11</span>—different types of protein particles; and <span class="html-italic">12</span>—surface passivation layer. Geometrical proportions of the figures are not observed for reasons of visual clarity (e.g. quartz crystal must be much thicker; many tens of periods of protrusions must fit on the area of the top electrode; particles must be smaller in reality).</p> Full article ">Figure 2
<p>Model development of quartz resonator with planar (<b>a</b>,<b>c</b>,<b>e</b>,<b>g</b>,<b>i</b>) and structured (<b>b</b>,<b>d</b>,<b>f</b>,<b>h</b>,<b>j</b>) top surfaces with direct solid/liquid contact (<b>a</b>,<b>b</b>) or coated with a thin surface passivation layer (<b>c</b>–<b>j</b>) and loaded by liquid in different cases: (<b>a</b>,<b>b</b>) Homogeneous liquid; (<b>c</b>,<b>d</b>) liquid with protein particles of one type that are evenly distributed throughout the liquid volume; (<b>e</b>,<b>f</b>) liquid with charged particles of one type that are concentrated near the quartz surface due to an external electric field; (<b>g</b>,<b>h</b>) liquid with several types of particles separated in external electric field made by positively charged additional top electrode, which concentrate positively charged particles near the quartz surface and sorts out negatively charged particles; (<b>i</b>,<b>j</b>) liquid with several types of particles separated in external electric field made by negatively charged additional top electrode, which concentrates negatively charged particles near the quartz surface and sorts out positively charged particles. Schematic model of sensor structure consists of measured liquid (1); quartz crystal resonator (2) of <span class="html-italic">H</span> thickness; bottom electrode (3) under alternating probing potential; top grounded electrode (4); set of protrusions (5) of <span class="html-italic">h</span> height, <span class="html-italic">a</span> distance and <span class="html-italic">b</span> width, placed on the top electrode surface; surface passivation layer (6); positively (7) and negatively (8, 9) charged protein particles of different mass and morphology; addition electrode (10) attached to the <span class="html-italic">L</span> height ceiling of the microfluidic channel (11) and charged (12). Geometrical proportions of the figures are not observed for reasons of visual clarity (e.g. particles must be smaller in reality).</p> Full article ">Figure 3
<p>Resonant frequency of a quartz crystal sensor depending on the method of structuring its surface at various filling factors and an average height of 200 nm (<b>a</b>); shift of the resonant frequency of a quartz resonator, depending on the method of structuring its surface at a filling factor of 20 % and different block heights (<b>b</b>); resonator frequency shift depending on liquid density for different heights of structural elements (<b>c</b>); frequency shift of a quartz resonator with structured surface elements of 1 and 3 µm height with a change in the liquid density and viscosity of 1 % (<b>d</b>). Subscript “0” means original liquid parameters.</p> Full article ">Figure 4
<p>Average concentration of proteins (shown by red and blue curves) trapped in the space between the barriers of the sensor with the structured surface at two different <span class="html-italic">pH</span> values (<b>a</b>) and the difference response from two sensors (with smooth and structured surfaces) in the form of a shift (shown by blue curves) of their resonance frequencies (<b>b</b>) by manipulation of the model buffer solution with proteins by external pulsed field (indicated by a black broken line).</p> Full article ">Figure 5
<p>The difference response from two sensors (with smooth and structured surfaces) in the form of a shift of their resonance frequencies to manipulation of a model buffer solution with proteins by an external pulsed field at different content of globulin (x = [ImG]/[Albumin], [Albumin] = 2 g/L) for pH = 6 (<b>a</b>) and pH = 9 (<b>b</b>). Frequency shifts (left axes) are shown with coloured curves, and pulse actions (right axes) are shown with a black broken line.</p> Full article ">Figure 6
<p>Average concentration of proteins trapped in the space between the barriers of the sensor with the structured surface at two different pH values (<b>a</b>) and the difference response from two sensors (with smooth and structured surfaces) in the form of a shift of their resonance frequencies (<b>b</b>) by manipulation of the model buffer solution with proteins by external pulsed field. Concentrations and frequency shifts (left axes) are shown with coloured curves, and pulse actions (right axes) are shown with a black broken line.</p> Full article ">Figure 7
<p>The difference response from two sensors (with smooth and structured surfaces) in the form of a shift of their resonance frequencies to manipulation of a model buffer solution with proteins by an external pulsed field at different ratios of oligomeric forms of uromodulin (y = [T&amp;HE (28)]/[T&amp;HE ( 7)], [T&amp;HE (28)] + [T&amp;HE (7)] = 120 mg/L) for a normal albumin level of 20 mg/L (<b>a</b>) and for its 3-fold increase (<b>b</b>). Frequency shifts (left axes) are shown with coloured curves, and pulse actions (right axes) are shown with a black broken line.</p> Full article ">
44 pages, 8038 KiB  
Review
The Key Role of Active Sites in the Development of Selective Metal Oxide Sensor Materials
by Artem Marikutsa, Marina Rumyantseva, Elizaveta A. Konstantinova and Alexander Gaskov
Sensors 2021, 21(7), 2554; https://doi.org/10.3390/s21072554 - 6 Apr 2021
Cited by 91 | Viewed by 7131
Abstract
Development of sensor materials based on metal oxide semiconductors (MOS) for selective gas sensors is challenging for the tasks of air quality monitoring, early fire detection, gas leaks search, breath analysis, etc. An extensive range of sensor materials has been elaborated, but no [...] Read more.
Development of sensor materials based on metal oxide semiconductors (MOS) for selective gas sensors is challenging for the tasks of air quality monitoring, early fire detection, gas leaks search, breath analysis, etc. An extensive range of sensor materials has been elaborated, but no consistent guidelines can be found for choosing a material composition targeting the selective detection of specific gases. Fundamental relations between material composition and sensing behavior have not been unambiguously established. In the present review, we summarize our recent works on the research of active sites and gas sensing behavior of n-type semiconductor metal oxides with different composition (simple oxides ZnO, In2O3, SnO2, WO3; mixed-metal oxides BaSnO3, Bi2WO6), and functionalized by catalytic noble metals (Ru, Pd, Au). The materials were variously characterized. The composition, metal-oxygen bonding, microstructure, active sites, sensing behavior, and interaction routes with gases (CO, NH3, SO2, VOC, NO2) were examined. The key role of active sites in determining the selectivity of sensor materials is substantiated. It was shown that the metal-oxygen bond energy of the MOS correlates with the surface acidity and the concentration of surface oxygen species and oxygen vacancies, which control the adsorption and redox conversion of analyte gas molecules. The effects of cations in mixed-metal oxides on the sensitivity and selectivity of BaSnO3 and Bi2WO6 to SO2 and VOCs, respectively, are rationalized. The determining role of catalytic noble metals in oxidation of reducing analyte gases and the impact of acid sites of MOS to gas adsorption are demonstrated. Full article
(This article belongs to the Special Issue Biennial State-of-the-Art Sensors Technology in Russia 2020-2021)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic of representation (<b>top</b>) of the surface of MOS possessing oxygen vacancies (V<sub>O</sub>) before (<b>a</b>) and after oxygen ionosorption (<b>b</b>); and the corresponding modulation of band energy levels (<b>bottom</b>): vacuum level (<span class="html-italic">E</span><sub>vac</sub>), conduction band bottom (<span class="html-italic">E<sub>C</sub></span>), Fermi level (<span class="html-italic">E<sub>F</sub></span>), donor states level (<span class="html-italic">E<sub>D</sub></span>), valence band top (<span class="html-italic">E<sub>V</sub></span>), potential energy surface barrier (−<span class="html-italic">eV</span><sub>S</sub>).</p> Full article ">Figure 2
<p>Schematic representation of the surface of <span class="html-italic">n</span>-type MOS with ionosorbed oxygen species (<b>b</b>) and after its interaction with reducing gas CO (<b>c</b>) and oxidizing gas NO<sub>2</sub> (<b>a</b>) within the chemisorption model of sensor response (<b>top</b>); and the corresponding modulation of band energy levels (<b>bottom</b>).</p> Full article ">Figure 3
<p>Schematic representation of the <span class="html-italic">n</span>-type MOS with oxygen vacancies and lattice oxygen anions (<b>b</b>), and after its interaction with reducing gas CO (<b>c</b>) and oxidizing gas NO<sub>2</sub> (<b>a</b>) within the oxygen vacancy model of sensor response (<b>top</b>); and the assumed changes in donor states population (<span class="html-italic">E</span><sub>D</sub>) and Fermi level positions (<span class="html-italic">E</span><sub>F</sub>) (<b>bottom</b>).</p> Full article ">Figure 4
<p>Positions of conduction band minima (<span class="html-italic">E</span><sub>c</sub>), valence band maxima (<span class="html-italic">E</span><sub>v</sub>) and electronegativity (χ) of <span class="html-italic">n</span>-type MOS, respective to vacuum level (<span class="html-italic">E</span><sub>vac</sub>). Adapted with permission using numeric data from ref. [<a href="#B54-sensors-21-02554" class="html-bibr">54</a>]. Copyright 2011 Elsevier. The oxides electronegativity corresponds to middle bandgap position. The levels of atomic orbitals for metal cations (M<sup>n+</sup> s<sup>0</sup>/d<sup>0</sup>) and oxygen anions (O<sup>2−</sup> 2p<sup>6</sup>) are shown schematically.</p> Full article ">Figure 5
<p>Types of active sites at the surface of tin oxide: noble metal cluster (NM), chemisorbed oxygen species (O<sub>2(ads)</sub>, O<sub>2</sub><sup>−</sup>, O<sup>−</sup>), charged oxygen vacancies (V<sub>O</sub><sup>−</sup>), partially reduced cations (Sn<sup>3+</sup>), coordinately unsaturated cations (Sn<sup>4+</sup><sub>cus</sub>), hydroxyl species (OH, Sn-OH), surface oxygen anions (O<sup>2−</sup>). Adapted with permission from ref. [<a href="#B79-sensors-21-02554" class="html-bibr">79</a>]. Copyright 2014 American Chemical Society.</p> Full article ">Figure 6
<p>Scheme of water molecule dissociative adsorption at metal oxide surface with the formation of bridging and terminal OH-groups. Reprinted with permission from ref. [<a href="#B84-sensors-21-02554" class="html-bibr">84</a>]. Copyright 2018 Springer.</p> Full article ">Figure 7
<p>FTIR spectra of nanocrystalline <span class="html-italic">n</span>-type MOS. Adapted with permission from ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B76-sensors-21-02554" class="html-bibr">76</a>] copyright 2019 Marikutsa, Rumyantseva, Gaskov, Batuk, Hadermann, Sarmadian, Saniz, Partoens and Lamoen Creative Commons Attribution License (CC BY); ref. [<a href="#B78-sensors-21-02554" class="html-bibr">78</a>] copyright 2021 Elsevier; ref. [<a href="#B79-sensors-21-02554" class="html-bibr">79</a>] copyright 2014 American Chemical Society.</p> Full article ">Figure 8
<p>XP-spectra of O 1s state for nanocrystalline <span class="html-italic">n</span>-type MOS. Adapted with permissions from ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B78-sensors-21-02554" class="html-bibr">78</a>] copyright 2021 Elsevier; [<a href="#B79-sensors-21-02554" class="html-bibr">79</a>] copyright 2014 American Chemical Society; ref. [<a href="#B87-sensors-21-02554" class="html-bibr">87</a>] copyright 2010 Elsevier.</p> Full article ">Figure 9
<p>HRTEM images of nanocrystalline ZnO, In<sub>2</sub>O<sub>3</sub>, and SnO<sub>2</sub>, and TEM image of WO<sub>3</sub>; the sample were annealed at 300 °C. The insets show electron diffraction patterns. Adapted with permissions from ref. [<a href="#B42-sensors-21-02554" class="html-bibr">42</a>] copyright 2019 by the authors (CC BY); ref. [<a href="#B76-sensors-21-02554" class="html-bibr">76</a>] copyright 2019 by the authors (CC BY); ref. [<a href="#B94-sensors-21-02554" class="html-bibr">94</a>] copyright 2015 Elsevier; ref. [<a href="#B95-sensors-21-02554" class="html-bibr">95</a>] copyright 2013 American Chemical Society.</p> Full article ">Figure 10
<p>SEM images of nanocrystalline TiO<sub>2</sub>, BaSnO<sub>3</sub>, and Bi<sub>2</sub>WO<sub>6</sub>. Adapted with permissions from ref. [<a href="#B63-sensors-21-02554" class="html-bibr">63</a>] copyright 2021 Elsevier; ref. [<a href="#B77-sensors-21-02554" class="html-bibr">77</a>] copyright 2015 by the authors (CC BY); ref. [<a href="#B96-sensors-21-02554" class="html-bibr">96</a>] copyright 2021 Springer Nature.</p> Full article ">Figure 11
<p>Temperature plots of ammonia desorption rate from the surface of <span class="html-italic">n</span>-type MOS (<b>a</b>). Adapted with permissions from ref. [<a href="#B63-sensors-21-02554" class="html-bibr">63</a>] copyright 2021 Elsevier; ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B78-sensors-21-02554" class="html-bibr">78</a>] copyright 2021 Elsevier; ref. [<a href="#B79-sensors-21-02554" class="html-bibr">79</a>] copyright 2014 American Chemical Society. TPD pattern of TiO<sub>2</sub> compared with mass-spectral (MS) analysis of desorbed gas (<b>b</b>).</p> Full article ">Figure 12
<p>Concentration of Brønsted and Lewis acid sites at the surface of nanocrystalline n-type MOS synthesized at 300 °C (ZnO, In<sub>2</sub>O<sub>3</sub>, SnO<sub>2</sub>, WO<sub>3</sub>) and 700 °C (TiO<sub>2</sub>) in relation to metal-oxygen bond energy. Adapted with permission from reference [<a href="#B78-sensors-21-02554" class="html-bibr">78</a>]. Copyright 2021 Elsevier.</p> Full article ">Figure 13
<p>EPR spectra of nanocrystalline <span class="html-italic">n</span>-type MOS. Adapted with permissions from ref. [<a href="#B79-sensors-21-02554" class="html-bibr">79</a>] copyright 2014 American Chemical Society; ref. [<a href="#B95-sensors-21-02554" class="html-bibr">95</a>] copyright 2013 American Chemical Society; ref. [<a href="#B96-sensors-21-02554" class="html-bibr">96</a>] copyright 2021 Springer Nature; ref. [<a href="#B101-sensors-21-02554" class="html-bibr">101</a>] copyright by the authors; ref. [<a href="#B102-sensors-21-02554" class="html-bibr">102</a>] copyright 2013 Elsevier.</p> Full article ">Figure 14
<p>Concentration of active sites at the surface of nanocrystalline <span class="html-italic">n</span>-type MOS in relation to metal-oxygen bond energy: oxidizing sites estimated from H<sub>2</sub> consumption in TPR at temperature below 300 °C (<b>a</b>), ionosorbed oxygen O<sub>2</sub><sup>−</sup> determined by EPR (<b>b</b>), and donor sites (V<sub>O</sub><sup>−</sup>) determined by EPR (<b>c</b>). The values are taken from <a href="#sensors-21-02554-t002" class="html-table">Table 2</a>.</p> Full article ">Figure 15
<p>Temperature plots of hydrogen consumption rate during TPR of nanocrystalline <span class="html-italic">n</span>-type MOS. Adapted with permissions from ref. [<a href="#B63-sensors-21-02554" class="html-bibr">63</a>] copyright 2021 Elsevier; ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B76-sensors-21-02554" class="html-bibr">76</a>] copyright by the authors (CC BY); ref. [<a href="#B78-sensors-21-02554" class="html-bibr">78</a>] copyright 2021 Elsevier; ref. [<a href="#B79-sensors-21-02554" class="html-bibr">79</a>] copyright 2014 American Chemical Society.</p> Full article ">Figure 16
<p>Unit cells of rutile-like tetragonal SnO<sub>2</sub>, perovskite-like cubic BaSnO<sub>3</sub> (<b>a</b>), monoclinic WO<sub>3</sub> and Aurivillius structure of Bi<sub>2</sub>WO<sub>6</sub> (<b>b</b>). Adapted with permissions from references [<a href="#B61-sensors-21-02554" class="html-bibr">61</a>,<a href="#B63-sensors-21-02554" class="html-bibr">63</a>,<a href="#B113-sensors-21-02554" class="html-bibr">113</a>]. Copyrights 2013 American Physical Society, 2021 Elsevier, 2014 Elsevier.</p> Full article ">Figure 17
<p>Sensitivity of pristine and RuO<sub>2</sub>-functionalized nanocrystalline <span class="html-italic">n</span>-type MOS to 20 ppm NH<sub>3</sub> at temperature 150–250 °C in relation to metal-oxygen bond energy in MOS (<b>a</b>) and acid sites concentration at the MOS surfaces (<b>b</b>). Adapted with permissions from ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B98-sensors-21-02554" class="html-bibr">98</a>] copyright 2018 Elsevier; copyright [<a href="#B116-sensors-21-02554" class="html-bibr">116</a>] 2019 John Wiley and sons; ref. [<a href="#B117-sensors-21-02554" class="html-bibr">117</a>] copyright 2012 Elsevier.</p> Full article ">Figure 18
<p>Sensitivity of pristine and PdO<sub>x</sub>-functionalized nanocrystalline <span class="html-italic">n</span>-type MOS to 20 ppm CO at temperature 250–300 °C in relation to metal-oxygen bond energy in MOS (<b>a</b>) and acid sites concentration at the MOS surfaces (<b>b</b>). Adapted with permissions from ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B76-sensors-21-02554" class="html-bibr">76</a>] copyright 2019 by the authors (BB BY); ref. [<a href="#B87-sensors-21-02554" class="html-bibr">87</a>] copyright 2010 Elsevier; ref. [<a href="#B97-sensors-21-02554" class="html-bibr">97</a>] copyright 2018 by the authors (CC BY); ref. [<a href="#B120-sensors-21-02554" class="html-bibr">120</a>] copyright 2015 by the authors (CC BY).</p> Full article ">Figure 19
<p>Sensitivity of pristine and Au-functionalized nanocrystalline <span class="html-italic">n</span>-type MOS to 20 ppm acetone (<b>a</b>,<b>c</b>) and 20 ppm methanol (<b>b</b>,<b>d</b>) at temperature 300 °C in relation to metal-oxygen bond energy in MOS (<b>a</b>,<b>b</b>) and sum of Lewis acid sites and oxidizing sites concentration at the MOS surfaces (<b>c</b>,<b>d</b>). Adapted with permission from ref. [<a href="#B78-sensors-21-02554" class="html-bibr">78</a>]. Copyright 2021 Elsevier.</p> Full article ">Figure 20
<p>Sensitivity of nanocrystalline <span class="html-italic">n</span>-type MOS to 1 ppm NO<sub>2</sub> at temperature 100–150 °C in relation to metal-oxygen bond energy (<b>a</b>) and donor sites concentration (<b>b</b>). Adapted with permissions from ref. [<a href="#B42-sensors-21-02554" class="html-bibr">42</a>] copyright 2019 by the authors (CC BY); ref. [<a href="#B63-sensors-21-02554" class="html-bibr">63</a>] copyright 2021 Elsevier; ref. [<a href="#B102-sensors-21-02554" class="html-bibr">102</a>] copyright 2013 Elsevier; ref. [<a href="#B123-sensors-21-02554" class="html-bibr">123</a>] copyright 2015 by the authors (CC BY).</p> Full article ">Figure 21
<p>Sensitivity of nanocrystalline SnO<sub>2</sub> and BaSnO<sub>3</sub> to NO<sub>2</sub> (2 ppm) at 100 °C, CO (50 ppm), NH<sub>3</sub> (20 ppm), H<sub>2</sub>S (2 ppm), H<sub>2</sub> (100 ppm), SO<sub>2</sub> (10 ppm), ethanol (20 ppm), methanol (20 ppm) at 300 °C (<b>a</b>) [<a href="#B77-sensors-21-02554" class="html-bibr">77</a>]; and sensitivity of nanocrystalline WO<sub>3</sub> and Bi<sub>2</sub>WO<sub>6</sub> to NO<sub>2</sub> (1 ppm) at 100 °C, ethanol (20 ppm) at 150 °C, SO<sub>2</sub> (2 ppm), formaldehyde (400 ppb), H<sub>2</sub>S (2 ppm) at 250 °C, CO (20 ppm), H<sub>2</sub> (50 ppm), NH<sub>3</sub> (20 ppm), acetone (2 ppm), benzene (2 ppm) at 300 °C (<b>b</b>) [<a href="#B63-sensors-21-02554" class="html-bibr">63</a>].</p> Full article ">Figure 22
<p>TEM micrographs, STEM images and EDX elemental maps of nanocrystalline SnO<sub>2</sub> and WO<sub>3</sub> functionalized by PdO<sub>x</sub> (<b>a</b>) and RuO<sub>2</sub> (<b>b</b>); and TEM micrographs with electron diffraction patterns of In<sub>2</sub>O<sub>3</sub> and TiO<sub>2</sub> functionalized by Au (<b>c</b>). Adapted with permissions from ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B78-sensors-21-02554" class="html-bibr">78</a>] copyright 2021 Elsevier; ref. [<a href="#B95-sensors-21-02554" class="html-bibr">95</a>] copyright 2013 American Chemical Society.</p> Full article ">Figure 23
<p>DRIFT spectra of pristine and functionalized by noble metal additives nanocrystalline SnO<sub>2</sub> (<b>a</b>) and WO<sub>3</sub> (<b>b</b>) exposed to CO (100 ppm (<b>a</b>); 200 ppm (<b>b</b>)) at room temperature for 1 h. Adapted with permissions from ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B118-sensors-21-02554" class="html-bibr">118</a>] copyright 2015 American Chemical Society.</p> Full article ">Figure 23 Cont.
<p>DRIFT spectra of pristine and functionalized by noble metal additives nanocrystalline SnO<sub>2</sub> (<b>a</b>) and WO<sub>3</sub> (<b>b</b>) exposed to CO (100 ppm (<b>a</b>); 200 ppm (<b>b</b>)) at room temperature for 1 h. Adapted with permissions from ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B118-sensors-21-02554" class="html-bibr">118</a>] copyright 2015 American Chemical Society.</p> Full article ">Figure 24
<p>DRIFT spectra of pristine and functionalized by noble metal additives nanocrystalline SnO<sub>2</sub> (<b>a</b>) and WO<sub>3</sub> (<b>b</b>) exposed to NH<sub>3</sub> (100 ppm (<b>a</b>); 200 ppm (<b>b</b>)) at 200 °C for 1 h. Adapted with permissions from ref. [<a href="#B75-sensors-21-02554" class="html-bibr">75</a>] copyright 2018 Elsevier; ref. [<a href="#B118-sensors-21-02554" class="html-bibr">118</a>] copyright 2015 American Chemical Society.</p> Full article ">Figure 25
<p>Concentration of oxidizing sites at the surface of pristine and Au-functionalized <span class="html-italic">n</span>-type MOS estimated from H<sub>2</sub> consumption in TPR at temperature below 300 °C (<b>a</b>) in comparison with the sensitivity to 20 ppm of acetone (<b>b</b>) and 20 ppm of methanol (<b>c</b>) as a function of metal-oxygen bond energy in MOS. Sensitivity of Au-functionalized MOS to 20 ppm of acetone (<b>d</b>) and 20 ppm of methanol (<b>e</b>) in relation to of oxidizing sites concentration (<b>d</b>,<b>e</b>). Operation temperature of sensors was 250–300 °C for pristine MOS and 150–225 °C for Au-functionalized MOS. Adapted with permission from ref. [<a href="#B78-sensors-21-02554" class="html-bibr">78</a>]. Copyright 2021 Elsevier.</p> Full article ">
27 pages, 18630 KiB  
Review
3D Sensors for Sewer Inspection: A Quantitative Review and Analysis
by Chris H. Bahnsen, Anders S. Johansen, Mark P. Philipsen, Jesper W. Henriksen, Kamal Nasrollahi and Thomas B. Moeslund
Sensors 2021, 21(7), 2553; https://doi.org/10.3390/s21072553 - 6 Apr 2021
Cited by 29 | Viewed by 6157
Abstract
Automating inspection of critical infrastructure such as sewer systems will help utilities optimize maintenance and replacement schedules. The current inspection process consists of manual reviews of video as an operator controls a sewer inspection vehicle remotely. The process is slow, labor-intensive, and expensive [...] Read more.
Automating inspection of critical infrastructure such as sewer systems will help utilities optimize maintenance and replacement schedules. The current inspection process consists of manual reviews of video as an operator controls a sewer inspection vehicle remotely. The process is slow, labor-intensive, and expensive and presents a huge potential for automation. With this work, we address a central component of the next generation of robotic inspection of sewers, namely the choice of 3D sensing technology. We investigate three prominent techniques for 3D vision: passive stereo, active stereo, and time-of-flight (ToF). The Realsense D435 camera is chosen as the representative of the first two techniques wheres the PMD CamBoard pico flexx represents ToF. The 3D reconstruction performance of the sensors is assessed in both a laboratory setup and in an outdoor above-ground setup. The acquired point clouds from the sensors are compared with reference 3D models using the cloud-to-mesh metric. The reconstruction performance of the sensors is tested with respect to different illuminance levels and different levels of water in the pipes. The results of the tests show that the ToF-based point cloud from the pico flexx is superior to the output of the active and passive stereo cameras. Full article
(This article belongs to the Special Issue Computer Vision for 3D Perception and Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Flow of the literature search. The records of the seven scholarly databases are manually screened based on the title, abstract, and keywords. In a subsequent stage, the records are combined and checked for duplicates. In the final process (light blue), the full-text of every record is manually assessed.</p> Full article ">Figure 2
<p>Different variations of depth sensing techniques.</p> Full article ">Figure 3
<p>Robotic setup for the laboratory and outdoor experiments. The setup is upgraded for the outdoor experiments with a new light source and a blue box that protects the Raspberry Pi from dirt.</p> Full article ">Figure 4
<p>Three different types of point cloud comparison. Red markings signify the reference, while green and gray signify the measured object.</p> Full article ">Figure 5
<p>Cloud-to-mesh distances from the depth sensor to the 3D model (not shown). Screenshots from CloudCompare [<a href="#B78-sensors-21-02553" class="html-bibr">78</a>]. Please note that the color scale vary between the plots.</p> Full article ">Figure 6
<p>Laboratory experimental setup.</p> Full article ">Figure 7
<p>Outdoor above-ground experimental setup. The four wells are laid out in a square configuration with 5m between each corner.</p> Full article ">Figure 8
<p>Water level setup. A piece of acrylic glass is inserted at both ends of the pipe in order to accommodate water.</p> Full article ">Figure 9
<p>Cross-sectional view of the sewer pipe. The water level is measured by <span class="html-italic">h</span> which is the line orthogonal to the water line and the bottom of the pipe.</p> Full article ">Figure 10
<p>View of the LEGO-based robot located in the laboratory (<b>a</b>) and outdoor (<b>b</b>) pipes. The lux meter is visible in the top of image (<b>a</b>). The lux meter was not present during the recording of sensor data.</p> Full article ">Figure 11
<p>Mean (<b>top</b>), standard deviation (<b>middle</b>) of the reconstruction error, and (<b>bottom</b>), average number of valid depth points from each depth image of the camera. C2M distance: cloud to mesh distance.</p> Full article ">Figure 12
<p>Illustration of the 3D to 2D mapping of the cylinder. The cross section of the cylinder is seen from the <span class="html-italic">XY</span>-plane. The angle <math display="inline"><semantics> <mi>θ</mi> </semantics></math> is mapped onto the <span class="html-italic">y</span>-axis of the two-dimensional plots.</p> Full article ">Figure 13
<p>2D projection of the cloud-to-mesh distance from the three sensor configurations in the laboratory setup. The color coding of the cloud represents the cloud-to-mesh distance. If points overlay each other on the 2D projection, the point with the largest cloud-to-mesh distance is plotted on top and is thus fully visible.</p> Full article ">Figure 14
<p>2D projection of the cloud-to-mesh distance from the three sensor configurations in the outdoor setup. The color coding of the cloud represents the cloud-to-mesh distance. If points overlay each other on the 2D projection, the point with the largest cloud-to-mesh distance is plotted on top and is thus fully visible.</p> Full article ">Figure 15
<p>Mean point density and reconstruction error (cloud-to-mesh distance) with respect to the depth (z) coordinate of the measurement.</p> Full article ">Figure 16
<p>Reconstructed point cloud for the water level experiments by the Realsense-On camera. The water level and pipe geometry is indicated with the orange line. The points are color-coded according to the cloud-to-mesh distance to the reference pipe. (<b>Top</b>) unfolded view of the reconstructed pipe. The water level is indicated with an orange dotted line. (<b>Middle</b>) View from the XY-plane of the reconstructed pipe. (<b>Bottom</b>) Zoomed in view of the XY-plane of the reconstructed pipe.</p> Full article ">Figure 17
<p>Reconstructed point cloud for the water level experiments by the pico flexx camera. The water level and pipe geometry is indicated with the orange line. The points are color-coded according to the cloud-to-mesh distance to the reference pipe. (<b>Top</b>) unfolded view of the reconstructed pipe. The water level is indicated with an orange dotted line. (<b>Middle</b>) View from the XY-plane of the reconstructed pipe. (<b>Bottom</b>) Zoomed in view of the XY-plane of the reconstructed pipe.</p> Full article ">
10 pages, 22287 KiB  
Communication
Leptospira interrogans Outer Membrane Protein-Based Nanohybrid Sensor for the Diagnosis of Leptospirosis
by Vivek Verma, Deepak Kala, Shagun Gupta, Harsh Kumar, Ankur Kaushal, Kamil Kuča, Natália Cruz-Martins and Dinesh Kumar
Sensors 2021, 21(7), 2552; https://doi.org/10.3390/s21072552 - 6 Apr 2021
Cited by 20 | Viewed by 4011
Abstract
Leptospirosis is an underestimated tropical disease caused by the pathogenic Leptospira species and responsible for several serious health problems. Here, we aimed to develop an ultrasensitive DNA biosensor for the rapid and on-site detection of the Loa22 gene of Leptospira interrogans using a [...] Read more.
Leptospirosis is an underestimated tropical disease caused by the pathogenic Leptospira species and responsible for several serious health problems. Here, we aimed to develop an ultrasensitive DNA biosensor for the rapid and on-site detection of the Loa22 gene of Leptospira interrogans using a gold nanoparticle–carbon nanofiber composite (AuN/CNF)-based screen-printed electrode. Cyclic voltammetry and electrochemical impedance were performed for electrochemical analysis. The sensitivity of the sensor was 5431.74 μA/cm2/ng with a LOD (detection limit) of 0.0077 ng/μL using cyclic voltammetry. The developed DNA biosensor was found highly specific to the Loa22 gene of L. interrogans, with a storage stability at 4 °C for 180 days and a 6% loss of the initial response. This DNA-based sensor only takes 30 min for rapid detection of the pathogen, with a higher specificity and sensitivity. The promising results obtained suggest the application of the developed sensor as a point of care device for the diagnosis of leptospirosis. Full article
(This article belongs to the Special Issue Screen-Printed Electrochemical Sensors and Their Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Characterization of the DNA sensor fabrication steps using a Fourier-transform infrared spectroscopy (FTIR) analysis of (<b>A</b>) the AuN/CNF bare electrode and (<b>B</b>) after modifications of the AuN/CNF electrode surface with a single stranded DNA (ssDNA) probe (AuN/CNF/ssDNA<sub>probe</sub>) at a frequency scan of 500–3500 cm<sup>−1</sup>.</p> Full article ">Figure 2
<p>Raman spectrum of (<b>A</b>) the bare AuN/CNF electrode and (<b>B</b>) 5′amino-linked ssDNA probe-modified AuN/CNF electrode (AuN/CNF/ssDNA<sub>(probe)</sub>).</p> Full article ">Figure 3
<p>Voltammetric analysis of the developed DNA sensor in different phases of the fabrication, including (A) the AuN/CNF electrode (bare), (B) AuN/CNF/ssDNA (probe), and (C–O) hybridization with single-stranded GDNA of <span class="html-italic">Leptospira interrogans</span>. The insert (<b>I</b>) shows a linear curve for the calculation of the limit of detection (LOD) and (<b>II</b>) shows a hyperbolic curve plotted between the relative peak current <span class="html-italic">Ip</span> with respect to probe with different concentrations of hybridizing ssGDNA of <span class="html-italic">L. interrogans</span>.</p> Full article ">Figure 4
<p>Comparison of the electrochemical impedance spectra of the DNA chip fabrication steps, including (a) the immobilization of the 5′amino-linked DNA probe and (b–h) hybridization with various concentrations of ssDNA of <span class="html-italic">L. interrogans</span> using 1-mM Potassium ferricyanide solution.</p> Full article ">Figure 5
<p>Cyclic voltametric analysis of DNA sensor selectivity using cDNA (complementary DNA) and a sequence with different numbers of mismatched bases. The insert shows the relative peak current values % <span class="html-italic">Ip</span> (with respect to the probe) of the DNA sensor with cDNA and different numbers of mismatched bases.</p> Full article ">Figure 6
<p>Evaluation of the DNA sensor specificity with <span class="html-italic">L. interrogans</span> and other bacterial species (<span class="html-italic">Escherichia coli, Staphylococcus aureus</span>, and <span class="html-italic">Klebsiella pneumoniae</span>) using cyclic voltametric studies (NC= Negative Control). The insert shows the relative <span class="html-italic">Ip</span> values (with respect to the probe) of the DNA sensor with the hybridizing GDNA of <span class="html-italic">L. interrogans</span> and other bacteria.</p> Full article ">Scheme 1
<p>Illustration of steps involved in the construction of the DNA sensor. AuN: Gold Nanoparticles, CNF: Carbon Nanofiber, SPE: Screen-Printed Electrode, MPA: 3-Mercaptopropionic acid, CV: Cyclic Voltammetry, EIS: Electrochemical Impedance Spectroscopy, EDC: 1-Ethyl-3-(3-dimethyl aminopropyl) carbodiimide, and NHS: N-Hydroxysuccinimide.</p> Full article ">
17 pages, 4791 KiB  
Article
An 18.8–33.9 GHz, 2.26 mW Current-Reuse Injection-Locked Frequency Divider for Radar Sensor Applications
by Kwang-Il Oh, Goo-Han Ko, Jeong-Geun Kim and Donghyun Baek
Sensors 2021, 21(7), 2551; https://doi.org/10.3390/s21072551 - 6 Apr 2021
Cited by 6 | Viewed by 2877
Abstract
An 18.8–33.9 GHz, 2.26 mW current-reuse (CR) injection-locked frequency divider (ILFD) for radar sensor applications is presented in this paper. A fourth-order resonator is designed using a transformer with a distributed inductor for wideband operating of the ILFD. The CR core is employed [...] Read more.
An 18.8–33.9 GHz, 2.26 mW current-reuse (CR) injection-locked frequency divider (ILFD) for radar sensor applications is presented in this paper. A fourth-order resonator is designed using a transformer with a distributed inductor for wideband operating of the ILFD. The CR core is employed to reduce the power consumption compared to conventional cross-coupled pair ILFDs. The targeted input center frequency is 24 GHz for radar application. The self-oscillated frequency of the proposed CR-ILFD is 14.08 GHz. The input frequency locking range is from 18.8 to 33.8 GHz (57%) at an injection power of 0 dBm without a capacitor bank or varactors. The proposed CR-ILFD consumes 2.26 mW of power from a 1 V supply voltage. The entire die size is 0.75 mm × 0.45 mm. This CR-ILFD is implemented in a 65 nm complementary metal-oxide semiconductor (CMOS) technology. Full article
(This article belongs to the Special Issue Advanced CMOS Integrated Circuit Design and Application)
►▼ Show Figures

Figure 1

Figure 1
<p>Conventional phase-locked loop with mm-Wave frequency divider.</p> Full article ">Figure 2
<p>(<b>a</b>) Schematic of the conventional cross-coupled pair ILFD with second-order resonator and (<b>b</b>) phasor diagram for the basic principle of the conventional ILFD.</p> Full article ">Figure 3
<p>Graphs of magnitude of load impedance against angular frequency; (<b>a</b>) normal case, (<b>b</b>) abnormal case.</p> Full article ">Figure 4
<p>Schematic of (<b>a</b>) conventional cross-coupled pair ILFD with fourth-order resonator and (<b>b</b>) CR core-based ILFD.</p> Full article ">Figure 5
<p>(<b>a</b>) Simulated magnitude plot and (<b>b</b>) phase plot of second-order resonator-based ILFD and fourth-order resonator-based ILFD.</p> Full article ">Figure 6
<p>Schematic of the proposed CR-ILFD.</p> Full article ">Figure 7
<p>(<b>a</b>) Simulated magnitude plot and (<b>b</b>) phase plot of the fourth-order resonator-based ILFD and proposed CR-ILFD.</p> Full article ">Figure 8
<p>(<b>a</b>) Modeling of the fourth-order resonator using a transformer with distributed inductor. (<b>a</b>) Modeling of including the parasitic capacitors and resistors. (<b>b</b>) Approximate modeling applied to simplify calculations.</p> Full article ">Figure 9
<p>Flowchart of the design approach for the proposed CR-ILFD.</p> Full article ">Figure 10
<p>Die photograph of the proposed CR-ILFD.</p> Full article ">Figure 11
<p>Measurement setup for the proposed CR-ILFD.</p> Full article ">Figure 12
<p>(<b>a</b>) Measured locking range results of the proposed CR-ILFD with different V<sub>inj,DC</sub>; (<b>b</b>) measured and simulated locking range results of the proposed CR-ILFD.</p> Full article ">Figure 13
<p>(<b>a</b>) Measured maximum and minimum operation frequency of the proposed CR-ILFD with different <span class="html-italic">V<sub>inj,DC</sub></span>; (<b>b</b>) Measured phase noise of input and output signal.</p> Full article ">Figure 14
<p>Spectrums of the output signal (<b>a</b>) when the proposed CR-ILFD self-oscillates; (<b>b</b>) when the proposed CR-ILFD is locked with a 28 GHz injection signal.</p> Full article ">Figure 15
<p>Full span spectrums (<b>a</b>) when the minimum input frequency is injected (18.8 GHz); (<b>b</b>) when the maximum input frequency is injected (33.8 GHz). The power difference between the output and input signals is approximately 10 dB or more.</p> Full article ">
13 pages, 296 KiB  
Article
Beam Allocation and Power Optimization for Energy-Efficiency in Multiuser mmWave Massive MIMO System
by Saidiwaerdi Maimaiti, Gang Chuai, Weidong Gao and Jinxi Zhang
Sensors 2021, 21(7), 2550; https://doi.org/10.3390/s21072550 - 6 Apr 2021
Cited by 3 | Viewed by 2372
Abstract
This paper studies beam allocation and power optimization scheme to decrease the hardware cost and downlink power consumption of a multiuser millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) system. Our target is to improve energy efficiency (EE) and decrease power consumption without obvious [...] Read more.
This paper studies beam allocation and power optimization scheme to decrease the hardware cost and downlink power consumption of a multiuser millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) system. Our target is to improve energy efficiency (EE) and decrease power consumption without obvious system performance loss. To this end, we propose a beam allocation and power optimization scheme. First, the problem of beam allocation and power optimization is formulated as a multivariate mixed-integer non-linear programming problem. Second, due to the non-convexity of this problem, we decompose it into two sub-problems which are beam allocation and power optimization. Finally, the beam allocation problem is solved by using a convex optimization technique. We solve the power optimization problem in two steps. First, the non-convex problem is converted into a convex problem by using a quadratic transformation scheme. The second step implements Lagrange dual and sub-gradient methods to solve the optimization problem. Performance analysis and simulation results show that the proposed algorithm performs almost identical to the exhaustive search (ES) method, while the greedy beam allocation and suboptimal beam allocation methods are far from the ES. Furthermore, experiment results demonstrated that our proposed algorithm outperforms the compared the greedy beam allocation method and the suboptimal beam allocation scheme in terms of average service ratio. Full article
(This article belongs to the Special Issue Energy-Efficient Resource Allocation for beyond 5G and IoT Systems)
►▼ Show Figures

Figure 1

Figure 1
<p>EE (energy efficiency) versus iterative with <math display="inline"><semantics> <msub> <mi>P</mi> <mi>t</mi> </msub> </semantics></math> = 30 dBm, <math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>2</mn> </mrow> </semantics></math>, <span class="html-italic">U</span> = 20.</p> Full article ">Figure 2
<p>EE versus transmit power with <math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>2</mn> </mrow> </semantics></math>, <span class="html-italic">U</span> = 20.</p> Full article ">Figure 3
<p>EE versus the number of users with <math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>2</mn> </mrow> </semantics></math>, <math display="inline"><semantics> <msub> <mi>P</mi> <mi>t</mi> </msub> </semantics></math> = 30 dBm.</p> Full article ">Figure 4
<p>Average service ratio versus total number of users <span class="html-italic">U</span> with <math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>2</mn> </mrow> </semantics></math>, <span class="html-italic">M</span> = 64, <math display="inline"><semantics> <msub> <mi>P</mi> <mi>t</mi> </msub> </semantics></math> = 30 dBm.</p> Full article ">Figure 5
<p>Average service ratio versus total number of antennas <span class="html-italic">M</span> with <math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>2</mn> </mrow> </semantics></math>, <span class="html-italic">U</span> = 10, <math display="inline"><semantics> <msub> <mi>P</mi> <mi>t</mi> </msub> </semantics></math> = 30 dBm.</p> Full article ">Figure 6
<p>EE versus the required minimum data rate <math display="inline"><semantics> <msubsup> <mi>R</mi> <mrow> <mi>m</mi> <mo>,</mo> <mi>u</mi> </mrow> <mo movablelimits="true" form="prefix">min</mo> </msubsup> </semantics></math> with <math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>2</mn> </mrow> </semantics></math>, <math display="inline"><semantics> <msub> <mi>P</mi> <mi>t</mi> </msub> </semantics></math> = 30 dBm.</p> Full article ">
10 pages, 3617 KiB  
Communication
Use of the Composite Properties of a Microwave Resonator to Enhance the Sensitivity of a Honey Moisture Sensor
by José R. Reyes-Ayona, Eloisa Gallegos-Arellano and Juan M. Sierra-Hernández
Sensors 2021, 21(7), 2549; https://doi.org/10.3390/s21072549 - 6 Apr 2021
Cited by 1 | Viewed by 2206
Abstract
A moisture sensor based on a composite resonator is used to measure different honey samples, which include imitation honey. The sensor changes its frequency response in accordance with the dielectric permittivity that it detects in the measured samples. Although reflectometry sensors have been [...] Read more.
A moisture sensor based on a composite resonator is used to measure different honey samples, which include imitation honey. The sensor changes its frequency response in accordance with the dielectric permittivity that it detects in the measured samples. Although reflectometry sensors have been used to measure the percentage of moisture in honey for almost a century, counterfeiters have achieved that their apocryphal product is capable of deceiving these kinds of sensors. Metamaterial features of the composite resonators are expected to improve their response when detecting lossy samples such as organic samples. It is also sought that these sensors manage to detect small differences not only in the real parts of the dielectric permitivities of samples but also in their imaginary parts, and, thus, the sensors are able to discern between real honey and slightly altered honey. Effectively, not only was it possible to improve the response of the sensors by using lossy samples but it was also possible to identify counterfeit honey. Full article
(This article belongs to the Special Issue Biomedical Microwave Sensors for Point-of-Care Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Top view of sensor layouts for circular and rectangular resonators. Orange microstrips are the feeding lines, grey lines are the external coupling lines, and light-grey lines are the resonant structures. Dots display short locations for sensors configuration <span class="html-italic">D</span><sub>1</sub>, <span class="html-italic">D</span><sub>2</sub>, and <span class="html-italic">D</span><sub>3</sub>. Colored arrow-lines are just illustrative to show the difference in distances seen by the traveling waves.</p> Full article ">Figure 2
<p>Reflection losses as a function of frequency of the rectangular resonator for three different configurations (<span class="html-italic">D</span><sub>1</sub>, <span class="html-italic">D</span><sub>2</sub>, and <span class="html-italic">D</span><sub>3</sub>). This frequency range shows the three first modes <span class="html-italic">N</span> of a resonator at ~0.5 GHz for the first mode, at ~1.5 GHz for the second mode, and at ~2.6 GHz for the third mode. Even though it is the same resonator and modes, reflection loss values are quite different.</p> Full article ">Figure 3
<p>Dispersion relations of the sensor for configurations <span class="html-italic">D</span><sub>1</sub> (blue solid line), <span class="html-italic">D</span><sub>2</sub> (red dashed line), and <span class="html-italic">D</span><sub>3</sub> (black dotted line). Blue x symbols, red ∆ symbols, and black + symbols are placed at the resonant frequency values for the three first modes of configuration <span class="html-italic">D</span><sub>1</sub>, <span class="html-italic">D</span><sub>2</sub>, and <span class="html-italic">D</span><sub>3</sub>, respectively. The dispersion relation reveals forward and backward propagation zones. Between 1.5 GHz and 3 GHz, there are several changes in the direction of propagation and some of them in a very abrupt way, in this zone, which is where the sensor might be more perceptible of small changes in its environment.</p> Full article ">Figure 4
<p>(<b>a</b>) Real part, and (<b>b</b>) imaginary part as a function of frequency of the input impedance of the sensor for configurations <span class="html-italic">D</span><sub>1</sub> (blue solid line), <span class="html-italic">D</span><sub>2</sub> (red dashed-dotted line), and <span class="html-italic">D</span><sub>3</sub> (black dotted line). Blue x symbols, red ∆ symbols, and black + symbols are placed at the resonant frequency values for the three first modes of configuration <span class="html-italic">D</span><sub>1</sub>, <span class="html-italic">D</span><sub>2</sub>, and <span class="html-italic">D</span><sub>3</sub>, respectively.</p> Full article ">Figure 5
<p>(<b>a</b>) A picture of some measured samples. For each of the six available honey varieties, eight samples were separated, weighted, and placed inside a plastic cup. (<b>b</b>) A picture of the measurement process where a sample is positioned on top of a sensor. The sensor is connected to a N9914A-21 Vector Network Analyzer (VNA) by means of a low loss precision test cable of the frequency response can be seen on the screen of the VNA.</p> Full article ">Figure 6
<p>Frequency response of a sensor for a single test where the sensor is configurated in a conventional zone for six honey samples. Frequency responses obtained for the different samples are almost identical, while differences are practically negligible. There is a variation in insertion losses of just 1 dB and in frequency resonance of only 6 MHz at most. In addition, the amplitudes of insertion losses are ~ −9.5 dB. Even though they are acceptable, they also are ordinary.</p> Full article ">Figure 7
<p>Frequency response of a sensor for a single test where the sensor is configurated in enhancing the performance zone for the same six honey samples used for <a href="#sensors-21-02549-f006" class="html-fig">Figure 6</a>. Two modes are displayed to show off the discrepancies. For the mode ~3.1 GHz, there is almost no change in the insertion loss values when honey samples are present or not. Nevertheless, for the mode of around 1.9 GHz, there is a significant change from −9 to −23 dB. The change in frequency when there is no sample and when a sample is placed on the sensor is consistent with the difference in relative permittivity of air and honey.</p> Full article ">Figure 8
<p>Frequency response of a 13 mm circular sensor for a single test where the sensor is configurated in a highly enhancing performance zone for the same six honey samples used for <a href="#sensors-21-02549-f006" class="html-fig">Figure 6</a> and <a href="#sensors-21-02549-f007" class="html-fig">Figure 7</a>. Two modes are once again displayed to show off the discrepancies between both modes. For the mode ~2.2 GHz presented on the left side, there is a change in the insertion loss values for No Sample (NS) and honey samples of around 10 dB. For the mode ~3.4 GHz presented on the right side, the difference in insertion loss values is at least of 17 dB for NS and honey samples. In addition, there is a significant discrepancy in the response for the imitation honey sample, which not only has a difference of 30 dB with NS, but also of more than 10 dB with other honey samples.</p> Full article ">
19 pages, 3962 KiB  
Article
Estimation of Tissue Attenuation from Ultrasonic B-Mode Images—Spectral-Log-Difference and Method-of-Moments Algorithms Compared
by Dinah Maria Brandner, Xiran Cai, Josquin Foiret, Katherine W. Ferrara and Bernhard G. Zagar
Sensors 2021, 21(7), 2548; https://doi.org/10.3390/s21072548 - 5 Apr 2021
Cited by 13 | Viewed by 4308
Abstract
We report on results from the comparison of two algorithms designed to estimate the attenuation coefficient from ultrasonic B-mode scans obtained from a numerical phantom simulating an ultrasound breast scan. It is well documented that this parameter significantly diverges between normal tissue and [...] Read more.
We report on results from the comparison of two algorithms designed to estimate the attenuation coefficient from ultrasonic B-mode scans obtained from a numerical phantom simulating an ultrasound breast scan. It is well documented that this parameter significantly diverges between normal tissue and malignant lesions. To improve the diagnostic accuracy it is of great importance to devise and test algorithms that facilitate the accurate, low variance and spatially resolved estimation of the tissue’s attenuation properties. A numerical phantom is realized using k-Wave, which is an open source Matlab toolbox for the time-domain simulation of acoustic wave fields that facilitates both linear and nonlinear wave propagation in homogeneous and heterogeneous tissue, as compared to strictly linear ultrasound simulation tools like Field II. k-Wave allows to simulate arbitrary distributions, resolved down to single voxel sizes, of parameters including the speed of sound, mass density, scattering strength and to include power law acoustic absorption necessary for simulation tasks in medical diagnostic ultrasound. We analyze the properties and the attainable accuracy of both the spectral-log-difference technique, and a statistical moments based approach and compare the results to known reference values from the sound field simulation. Full article
(This article belongs to the Special Issue Selected Papers from 6th International Conference on Sensors and Electronic Instrumentation Advances (SEIA’ 2020))
►▼ Show Figures

Figure 1

Figure 1
<p>Setup used for the measurement of tissue-mimicking phantoms. (<b>a</b>) Piston-type US transducer (ROHE-5604), (<b>b</b>) phantom, (<b>c</b>) needle hydrophone (HNA-0400, ONDA, Sunnyvale, CA, USA) with amplifier (AH-1100, ONDA, Sunnyvale, CA, USA). The thickness of the phantom is 2 cm.</p> Full article ">Figure 2
<p>Attenuation coefficient vs. concentration of aluminum oxide (grain size 0.3 μm) for homogeneously prepared tissue mimicking phantoms (as shown in <a href="#sensors-21-02548-f001" class="html-fig">Figure 1</a>b). Reprinted from ref. [<a href="#B15-sensors-21-02548" class="html-bibr">15</a>].</p> Full article ">Figure 3
<p>Ultrasonic wave’s envelope propagating downwards, focused at 35 mm and being scattered off discretely placed scatterers (indicated as red circles in exaggerated size). The chosen colormap for the log-compressed display extends from 100 kPa (orange) to 1 Pa (in blue).</p> Full article ">Figure 4
<p>Left diagram: decrease of sound pressure amplitude vs. depth along the center line within the volume in blue and the expected decrease in sound pressure for the assumed attenuation coefficient of 2.3 dB/(MHz cm) in red. Right diagram decrease of sound pressure level vs. depth estimated from a B-mode line (in dB ref. 100 kPa). Indicated in red is the expected decrease in amplitude vs. depth for the chosen attenuation of 2.3 dB/(MHz cm).</p> Full article ">Figure 5
<p>(<b>Left</b>): To define locations and indices within a sector scan used in the estimation algorithms. At the top the US transducer is assumed operating as a phased array. (<b>Right</b>): Simulated US B-mode image, exhibiting typical US speckles and shadowing behind a strongly absorbing region (a simulated lesion). Distal (bottom, subscript (...)<math display="inline"><semantics> <msub> <mrow/> <mi>d</mi> </msub> </semantics></math>) and proximal (top, subscript (...)<math display="inline"><semantics> <msub> <mrow/> <mi>p</mi> </msub> </semantics></math>) windows are indicated.</p> Full article ">Figure 6
<p>The blue graph represents the power spectral density for the ultrasound propagation in pure water, the red line is the measured spectrum obtained over the thickness of 2 cm of the tissue mimicking phantom depicted in <a href="#sensors-21-02548-f001" class="html-fig">Figure 1</a> which was designed to result in an <math display="inline"><semantics> <msub> <mi>α</mi> <mn>0</mn> </msub> </semantics></math> of 0.6 dB/(MHz cm), and the green line is the spectral-log-difference. The two estimators used, led to an estimated attenuation of: <math display="inline"><semantics> <mrow> <msub> <mover accent="true"> <mi>α</mi> <mo stretchy="false">^</mo> </mover> <mn>0</mn> </msub> <msub> <mrow> <mo>|</mo> </mrow> <mrow> <mi>S</mi> <mi>L</mi> <mi>D</mi> </mrow> </msub> </mrow> </semantics></math> = 0.5663 dB/(MHz cm) and <math display="inline"><semantics> <mrow> <msub> <mover accent="true"> <mi>α</mi> <mo stretchy="false">^</mo> </mover> <mn>0</mn> </msub> <msub> <mrow> <mo>|</mo> </mrow> <mrow> <mi>M</mi> <mi>o</mi> <mi>M</mi> </mrow> </msub> </mrow> </semantics></math> = 0.5759 dB/(MHz cm).</p> Full article ">Figure 7
<p>Comparison of the performance of both estimators. (<b>a</b>) The imaging geometry and the spectral density analysis windows. As a steering angle of <math display="inline"><semantics> <mrow> <mo>±</mo> <msup> <mn>15</mn> <mo>∘</mo> </msup> </mrow> </semantics></math> would appear very slim, the image was stretched to square, leading to speckles that appear wider than those seen in real ultrasound images. Furthermore, the distal and proximal windows are displayed larger than they actually are, to give an idea of what is done here, meaning there is not as much averaging done than could be assumed by inspecting the graphic. (<b>b</b>) Shows the simulated B-mode image rendered with correct time-gain-compensation. This raw data is analyzed by both attenuation estimating algorithms. (<b>c</b>) The color-coded overlaid estimation result for the spectral-log-difference algorithm. (<b>d</b>) Estimation result for the method-of-moments. (<b>e</b>) Probability density function for the spectral-log-difference algorithm. (<b>f</b>) Probability density function for the method-of-moments.</p> Full article ">Figure 8
<p>Result of estimating <math display="inline"><semantics> <msub> <mover accent="true"> <mi>a</mi> <mo stretchy="false">^</mo> </mover> <mn>2</mn> </msub> </semantics></math> from simulations of the homogeneous phantom (<b>left</b>). The true value is <math display="inline"><semantics> <mrow> <msub> <mi>a</mi> <mn>2</mn> </msub> <mo>=</mo> <mn>0.2</mn> </mrow> </semantics></math>, and the histogram for <math display="inline"><semantics> <msub> <mover accent="true"> <mi>a</mi> <mo stretchy="false">^</mo> </mover> <mn>2</mn> </msub> </semantics></math> (<b>right</b>). One can observe a rather large spread thus rendering the estimate to be of limited use in diagnosis.</p> Full article ">Figure 9
<p>(<b>a</b>) B-mode images of an inclusion with a radius of 7.5 mm positioned centered at a depth of 20 mm, left, the inclusion is mimicking a cyst (<math display="inline"><semantics> <mrow> <msub> <mi>α</mi> <mn>0</mn> </msub> <mo>=</mo> <mn>0.2</mn> </mrow> </semantics></math> dB/(MHz cm)). (<b>b</b>) B-mode images of an inclusion with a radius of 7.5 mm positioned centered at a depth of 20 mm, the inclusion is mimicking a malignant lesion (<math display="inline"><semantics> <mrow> <msub> <mi>α</mi> <mn>0</mn> </msub> <mo>=</mo> <mn>2.5</mn> </mrow> </semantics></math> dB/(MHz cm)). (<b>c</b>,<b>d</b>) Estimation results obtained from the application of the spectral-log-difference algorithm. (<b>e</b>,<b>f</b>) Estimation results obtained from the method-of-moments.</p> Full article ">
23 pages, 4617 KiB  
Article
Multi-Sensor Fault Detection, Identification, Isolation and Health Forecasting for Autonomous Vehicles
by Saeid Safavi, Mohammad Amin Safavi, Hossein Hamid and Saber Fallah
Sensors 2021, 21(7), 2547; https://doi.org/10.3390/s21072547 - 5 Apr 2021
Cited by 54 | Viewed by 10735
Abstract
The primary focus of autonomous driving research is to improve driving accuracy and reliability. While great progress has been made, state-of-the-art algorithms still fail at times and some of these failures are due to the faults in sensors. Such failures may have fatal [...] Read more.
The primary focus of autonomous driving research is to improve driving accuracy and reliability. While great progress has been made, state-of-the-art algorithms still fail at times and some of these failures are due to the faults in sensors. Such failures may have fatal consequences. It therefore is important that automated cars foresee problems ahead as early as possible. By using real-world data and artificial injection of different types of sensor faults to the healthy signals, data models can be trained using machine learning techniques. This paper proposes a novel fault detection, isolation, identification and prediction (based on detection) architecture for multi-fault in multi-sensor systems, such as autonomous vehicles.Our detection, identification and isolation platform uses two distinct and efficient deep neural network architectures and obtained very impressive performance. Utilizing the sensor fault detection system’s output, we then introduce our health index measure and use it to train the health index forecasting network. Full article
(This article belongs to the Special Issue Artificial Intelligence and Internet of Things in Autonomous Vehicles)
►▼ Show Figures

Figure 1

Figure 1
<p>Different forecasting categories for various applications [<a href="#B12-sensors-21-02547" class="html-bibr">12</a>].</p> Full article ">Figure 2
<p>Three vehicle bus data-points are depicted over time. The top figure shows the output of the Accelerator Pedal (AccP) sensor, the middle figure is for the Steering Wheel Angle (SWA) and the bottom figure is showing the output of the Brake Pressure (BP) sensor.</p> Full article ">Figure 3
<p>Examples of sensor faults. Data have been taken from the Audi autonomous driving dataset (A2D2). All four figures are faulty variants of a sample signal from the AccP sensor.</p> Full article ">Figure 4
<p>Proposed system architecture.</p> Full article ">Figure 5
<p>The convolution structures of the suggested 1D CNN adaptive configuration [<a href="#B26-sensors-21-02547" class="html-bibr">26</a>], e.g., “<math display="inline"><semantics> <mrow> <mn>9</mn> <mo>×</mo> <mn>1</mn> </mrow> </semantics></math> conv, 60, /1” indicates a convolutional layer with 60 kernels, size <math display="inline"><semantics> <mrow> <mn>9</mn> <mo>×</mo> <mn>1</mn> </mrow> </semantics></math>, stride 1; a box labeled “<math display="inline"><semantics> <mrow> <mn>4</mn> <mo>×</mo> <mn>1</mn> </mrow> </semantics></math> pool, /4” implies a max pooling layer with a size of <math display="inline"><semantics> <mrow> <mn>4</mn> <mo>×</mo> <mn>1</mn> </mrow> </semantics></math>, stride 4; ‘fc 20’ is a fully connected layer with the output dimension of 20. The numbers across boxes are the prior layers’ output sizes.</p> Full article ">Figure 6
<p>The proposed multi-class deep neural network (DNN) architecture in this study for fault isolation and identification.</p> Full article ">Figure 7
<p>Three different degradation scenarios for erratic, drift and spike fault types of the AccP sensor. (<b>a</b>) Steps taken to construct the HI from a linearly degraded AccP data stream with erratic fault. (<b>b</b>) Steps taken to construct the HI from an exponentially degraded AccP data stream with drift fault. (<b>c</b>) Steps taken to construct the HI from a sinusoidal degraded AccP data stream with spike fault. (<b>d</b>) Steps taken to construct the HI from a healthy AccP data stream.</p> Full article ">Figure 8
<p>Steps taken to construct the HI from a linearly degraded AccP data stream.</p> Full article ">Figure 9
<p>Steps taken to construct the health index (HI) from an exponentially degraded AccP data stream.</p> Full article ">Figure 10
<p>Steps taken to construct the HI from a sinusoidal degraded AccP data stream.</p> Full article ">Figure 11
<p>Sketch of static covariate multi-horizon forecasting and past-observed inputs.</p> Full article ">Figure 12
<p>Temporal Fusion Transformers (TFT) architectural design. TFT inputs static attributes, time-varying historical inputs and time-varying a priori known future inputs. Variable filtering is used to allow a judicious selection of the most excellent features based on input. Gated Residual Network Blocks allow efficient information flow with skip and gating layer connections. Time-dependent processing is built on LSTMs for local processing and multi-head attention for the aggregation of information at any time.</p> Full article ">Figure 13
<p>The confusion matrix of our detection system which achieved an accuracy of 99.84%.</p> Full article ">Figure 14
<p>Test data vs. TFT predictions.</p> Full article ">Figure 15
<p>Test data vs. TFT forecasts for three different quantiles.</p> Full article ">
16 pages, 7467 KiB  
Article
Non-Destructive and Quantitative Evaluation of Rebar Corrosion by a Vibro-Doppler Radar Method
by Takashi Miwa
Sensors 2021, 21(7), 2546; https://doi.org/10.3390/s21072546 - 5 Apr 2021
Cited by 11 | Viewed by 3709
Abstract
It is well known that evaluation of rebar corrosion is important for the maintenance of reinforced concrete structures, but, it is difficult to simply, quickly and quantitatively evaluate the amount of corrosion of rebars embedded in concrete by conventional non-destructive evaluation (NDE) methods [...] Read more.
It is well known that evaluation of rebar corrosion is important for the maintenance of reinforced concrete structures, but, it is difficult to simply, quickly and quantitatively evaluate the amount of corrosion of rebars embedded in concrete by conventional non-destructive evaluation (NDE) methods such as electrical, electromagnetic and mechanical method. This paper proposes a vibro-Doppler radar (VDR) measurement method to quantitatively evaluate rebar corrosion by measuring the vibration ability of the rebar forcibly vibrated in concrete by an excitation coil. It is experimentally demonstrated in RC test pieces that the rebar vibration displacement obtained by developed VDR method is valid and is less affected by the moisture in the concrete. In addition, simultaneous monitoring of the rebar vibration displacement of the test pieces is performed through an electrolytic corrosion test and the measured vibration displacement is compared to the rebar corrosion loss evaluated. As the results, it is cleared that the rebar vibration displacement starts to increase from slightly before the occurrences of corrosion crack on the concrete surface as the corrosion loss increases. It is also shown that the rebar vibration displacement becomes 4 times higher than that in initial condition at the rebar corrosion loss of 250 mg/cm2. This implies that the VDR has potential to nondestructively and quantitatively evaluate rebar corrosion in concrete. Full article
(This article belongs to the Special Issue Sensing Advancement and Health Monitoring of Transport Structures)
►▼ Show Figures

Figure 1

Figure 1
<p>Concept of vibro-Doppler radar.</p> Full article ">Figure 2
<p>Geometry of problem.</p> Full article ">Figure 3
<p>Block diagram of vibro-Doppler radar system based on network analyzer: (<b>a</b>) for non-modulation component; (<b>b</b>) for Doppler modulation component; (<b>c</b>) Photo of developed vibro-Doppler radar measurement system.</p> Full article ">Figure 4
<p>Rebar vibration characteristics in air: (<b>a</b>) Measurement setup; (<b>b</b>) Displacement spectrum obtained by a laser displacement sensor (LDS).</p> Full article ">Figure 5
<p>Measurement of rebar vibration displacement in concrete: (<b>a</b>) Experimental setup; (<b>b</b>) Displacement spectrum measured at the end of the rebar.</p> Full article ">Figure 6
<p>Measurement setup for VDR in concrete: (<b>a</b>) Antenna arrangement (<b>b</b>) Configuration of used antenna.</p> Full article ">Figure 7
<p>Results of Vibro-Doppler radar (VDR) measurement in concrete: (<b>a</b>) Envelope of VDR waveforms obtained for initial condition of test piece; (<b>b</b>) Envelope of VDR waveforms after immersion of test piece in water; (<b>c</b>) Vibration displacement of the rebar embedded in concrete.</p> Full article ">Figure 8
<p>Overview of experiment of: (<b>a</b>) Fabricated RC test piece; (<b>b</b>) Electrolytic corrosion test.</p> Full article ">Figure 9
<p>Protocol of the electrolytic corrosion test.</p> Full article ">Figure 10
<p>Cross-sectional photos of test pieces after terminating the electrolytic corrosion test.</p> Full article ">Figure 11
<p>Photographs of corrosion situation of the rebar taken out from the test pieces: (<b>a</b>) before and; (<b>b</b>) after removing the rust from the rebar.</p> Full article ">Figure 12
<p>Relationship between the corrosion loss of the rebar and the cumulative current in the electrolytic corrosion test.</p> Full article ">Figure 13
<p>Monitoring results obtained by VDR while the electrolytic corrosion test in the case of D16 rebar for: (<b>a</b>) non-modulation components; (<b>b</b>) Doppler modulation components. The enveloped waveforms are aligned in the lateral direction as a function of the cumulative current with the amplitude representing in color scale. The amplitude of the Doppler component is multiplied by 1000.</p> Full article ">Figure 14
<p>Rebar vibration displacement measured by VDR in the electrolytic corrosion test for: (<b>a</b>) D16 rebar; and (<b>b</b>) D22 rebar. The markers indicated by a circle, rectangle, and triangle mean the timing at occurrence of cracks on the top surface, that on the side surfaces of the concrete, and leakage of rust water from the crack, respectively.</p> Full article ">Figure 15
<p>Relationship between rebar corrosion loss and rebar vibration displacement obtained by VDR. The grey dotted lines mean standard deviation.</p> Full article ">
17 pages, 21229 KiB  
Article
A Compensation Method for Nonlinear Vibration of Silicon-Micro Resonant Sensor
by Yan Li, Hao Li, Yifeng Xiao, Le Cao and Zhan-She Guo
Sensors 2021, 21(7), 2545; https://doi.org/10.3390/s21072545 - 5 Apr 2021
Cited by 9 | Viewed by 3086
Abstract
A compensation method for nonlinear vibration of a silicon micro resonant sensor is proposed and evaluated to be effective through simulation and experimental analysis. Firstly, the parameter characterization model of the silicon micro resonant sensor is established, which presents significant nonlinearity because of [...] Read more.
A compensation method for nonlinear vibration of a silicon micro resonant sensor is proposed and evaluated to be effective through simulation and experimental analysis. Firstly, the parameter characterization model of the silicon micro resonant sensor is established, which presents significant nonlinearity because of the nonlinear vibration of the resonant beam. A verification circuit is devised to imitate the nonlinear behavior of the model by matching the simulation measurement error of the frequency offset produced by the circuit block with the theoretical counterparts obtained from the model. Secondly, the principle of measurement error compensation is studied, and the compensation method dealing with nonlinear characteristics of the resonant beam is proposed by introducing a compensation beam and corresponding differential operations. The measurement error, compensation rate, and measurement residual between the two scenarios that use single beam and double beams, respectively, are derived and are compared with their simulation and experimental counterparts. The results coincide with the predicted trend, which verifies the effectiveness of the compensation method. Full article
(This article belongs to the Special Issue MEMS Actuators and Sensors 2022)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Schematic diagram of the silicon-micro resonant pressure sensor.</p> Full article ">Figure 2
<p>Schematic diagram of micro element stress for the resonant beam.</p> Full article ">Figure 3
<p>Equivalent verification circuit block diagram of the silicon-micro resonant pressure sensor.</p> Full article ">Figure 4
<p>Equivalent verification circuit diagram of the silicon-micro resonant pressure sensor.</p> Full article ">Figure 5
<p>The frequency offset of the nonlinear coefficient.</p> Full article ">Figure 6
<p>The influence of nonlinear coefficient on measurement error.</p> Full article ">Figure 7
<p>Double silicon-micro resonant pressure sensor.</p> Full article ">Figure 8
<p>The block diagram of the error compensation system of the pressure sensor with the double resonant beam.</p> Full article ">Figure 9
<p>Measurement error of resonant pressure sensor caused by nonlinear vibration.</p> Full article ">Figure 10
<p>Nonlinear vibration frequency response test system.</p> Full article ">Figure 11
<p>Compensation rate of resonant pressure sensor caused by nonlinear vibration.</p> Full article ">Figure 12
<p>Measurement residual of the single resonant beam pressure sensor.</p> Full article ">Figure 13
<p>Measurement residual of the double resonant beam pressure sensor.</p> Full article ">
20 pages, 851 KiB  
Article
Performance Study of Distance-Weighting Approach with Loopy Sum-Product Algorithm for Multi-Object Tracking in Clutter
by Pranav U. Damale, Edwin K. P. Chong and Tian J. Ma
Sensors 2021, 21(7), 2544; https://doi.org/10.3390/s21072544 - 5 Apr 2021
Cited by 3 | Viewed by 2648
Abstract
In this paper, we explore the performance of the distance-weighting probabilistic data association (DWPDA) approach in conjunction with the loopy sum-product algorithm (LSPA) for tracking multiple objects in clutter. First, we discuss the problem of data association (DA), which is to infer the [...] Read more.
In this paper, we explore the performance of the distance-weighting probabilistic data association (DWPDA) approach in conjunction with the loopy sum-product algorithm (LSPA) for tracking multiple objects in clutter. First, we discuss the problem of data association (DA), which is to infer the correspondence between targets and measurements. DA plays an important role when tracking multiple targets using measurements of uncertain origin. Second, we describe three methods of data association: probabilistic data association (PDA), joint probabilistic data association (JPDA), and LSPA. We then apply these three DA methods for tracking multiple crossing targets in cluttered environments, e.g., radar detection with false alarms and missed detections. We are interested in two performance metrics: tracking accuracy and computation time. LSPA is known to be superior to PDA in terms of the former and to dominate JPDA in terms of the latter. Last, we consider an additional DA method that is a modification of PDA by incorporating a weighting scheme based on distances between position estimates and measurements. This distance-weighting approach, when combined with PDA, has been shown to enhance the tracking accuracy of PDA without significant change in the computation burden. Since PDA constitutes a crucial building block of LSPA, we hypothesize that DWPDA, when integrated with LSPA, would perform better under the two performance metrics above. Contrary to expectations, the distance-weighting approach does not enhance the performance of LSPA, whether in terms of tracking accuracy or computation time. Full article
(This article belongs to the Special Issue Multi-Sensor Fusion for Object Detection and Tracking)
►▼ Show Figures

Figure 1

Figure 1
<p>Bipartite graphical model formulation for data association at time <span class="html-italic">k</span>. The value assigned to <math display="inline"><semantics> <mrow> <msub> <mi>a</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> </mrow> </semantics></math> is an index to the measurement with which target <span class="html-italic">i</span> is hypothesized to be associated at time <span class="html-italic">k</span> and the value assigned to <math display="inline"><semantics> <mrow> <msub> <mi>b</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> </mrow> </semantics></math> is an index to the target with which measurement <span class="html-italic">j</span> is hypothesized to be associated at time <span class="html-italic">k</span>.</p> Full article ">Figure 2
<p>Factor graph representing the factorization of the joint posterior probability density function (PDF) <math display="inline"><semantics> <mrow> <mi>f</mi> <mo>(</mo> <msub> <mi>x</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>x</mi> <mn>2</mn> </msub> <mo>,</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>a</mi> <mn>2</mn> </msub> <mo>,</mo> <msub> <mi>b</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>b</mi> <mn>2</mn> </msub> <mo>|</mo> <msub> <mi>z</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>z</mi> <mn>2</mn> </msub> <mo>)</mo> </mrow> </semantics></math> according to Equation (<a href="#FD41-sensors-21-02544" class="html-disp-formula">41</a>), depicted for one time step. For simplicity, the time index <span class="html-italic">k</span> is omitted.</p> Full article ">Figure 3
<p>True target positions for three crossing targets with different clutter densities. (<b>a</b>) Clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>1</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>; and (<b>b</b>) clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>5</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>.</p> Full article ">Figure 4
<p>RMS position error for three crossing targets using DWPDA and LSPA with different clutter densities. (<b>a</b>) Clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>1</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>; and (<b>b</b>) clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>5</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>.</p> Full article ">Figure 5
<p>RMS position error for six crossing targets using DWPDA and LSPA with different clutter densities. (<b>a</b>) Clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>1</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>; and (<b>b</b>) clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>5</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>.</p> Full article ">Figure 6
<p>Average computation time for obtaining association probabilities using DWPDA and LSPA for tracking multiple crossing targets with different clutter densities. (<b>a</b>) Clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>1</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>; and (<b>b</b>) clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>5</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>. Error bars indicate 95% confidence intervals.</p> Full article ">Figure 7
<p>Tracking multiple crossing targets using LSPA and JPDA with clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>1</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>. (<b>a</b>) Average computation time for obtaining association probabilities; (<b>b</b>) average RMS position error. Error bars indicate 95% confidence intervals.</p> Full article ">Figure 8
<p>Tracking multiple crossing targets using loopy sum-product algorithm (LSPA) and joint probabilistic data association (JPDA) with clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>5</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>. (<b>a</b>) Average computation time for obtaining association probabilities; (<b>b</b>) average root mean square (RMS) position error. Error bars indicate 95% confidence intervals.</p> Full article ">Figure 9
<p>Tracking three crossing targets using LSPA and JPDA. (<b>a</b>) Average computation time for obtaining association probabilities; (<b>b</b>) average RMS position error. Error bars indicate 95% confidence intervals.</p> Full article ">Figure 10
<p>Tracking multiple crossing targets using LSPA and DWLSPA with clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>1</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>. (<b>a</b>) Average computation time for obtaining association probabilities; (<b>b</b>) average RMS position error. Error bars indicate 95% confidence intervals.</p> Full article ">Figure 11
<p>Tracking multiple crossing targets using LSPA and DWLSPA with clutter density <math display="inline"><semantics> <mrow> <mi>λ</mi> <mo>=</mo> <mn>1</mn> <mo>×</mo> <msup> <mn>10</mn> <mrow> <mo>−</mo> <mn>4</mn> </mrow> </msup> <mo>/</mo> </mrow> </semantics></math>m<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>.(<b>a</b>) Average computation time for obtaining association probabilities; (<b>b</b>) average RMS position error. Error bars indicate 95% confidence intervals.</p> Full article ">Figure 12
<p>Tracking three crossing targets using LSPA and DWLSPA. (<b>a</b>) Average computation time for obtaining association probabilities; (<b>b</b>) average RMS position error. Error bars indicate 95% confidence intervals.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-319
Previous Issue
Next Issue
Back to TopTop