CN111721336A - Self-interference micro-ring resonant cavity sensing classification identification method based on supervised learning - Google Patents
Self-interference micro-ring resonant cavity sensing classification identification method based on supervised learning Download PDFInfo
- Publication number
- CN111721336A CN111721336A CN202010156324.XA CN202010156324A CN111721336A CN 111721336 A CN111721336 A CN 111721336A CN 202010156324 A CN202010156324 A CN 202010156324A CN 111721336 A CN111721336 A CN 111721336A
- Authority
- CN
- China
- Prior art keywords
- micro
- training
- neural network
- waveguide
- sensing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000012549 training Methods 0.000 claims abstract description 71
- 230000005540 biological transmission Effects 0.000 claims abstract description 61
- 238000013528 artificial neural network Methods 0.000 claims abstract description 60
- 230000008033 biological extinction Effects 0.000 claims abstract description 59
- 238000001514 detection method Methods 0.000 claims abstract description 51
- 238000012360 testing method Methods 0.000 claims abstract description 30
- 239000013076 target substance Substances 0.000 claims abstract description 23
- 239000000126 substance Substances 0.000 claims abstract description 15
- 230000003287 optical effect Effects 0.000 claims description 28
- 239000000463 material Substances 0.000 claims description 22
- 230000008859 change Effects 0.000 claims description 14
- 238000001228 spectrum Methods 0.000 claims description 14
- 230000008878 coupling Effects 0.000 claims description 12
- 238000010168 coupling process Methods 0.000 claims description 12
- 238000005859 coupling reaction Methods 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 10
- 230000003993 interaction Effects 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 238000005259 measurement Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 8
- 238000013507 mapping Methods 0.000 claims description 7
- 238000000411 transmission spectrum Methods 0.000 claims description 6
- 230000001808 coupling effect Effects 0.000 claims 1
- 230000015572 biosynthetic process Effects 0.000 abstract 1
- 239000007789 gas Substances 0.000 description 45
- 239000000523 sample Substances 0.000 description 8
- 238000004088 simulation Methods 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 5
- 230000035945 sensitivity Effects 0.000 description 4
- 238000005253 cladding Methods 0.000 description 3
- 238000000295 emission spectrum Methods 0.000 description 3
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 2
- 239000006185 dispersion Substances 0.000 description 2
- 239000002105 nanoparticle Substances 0.000 description 2
- 229920000642 polymer Polymers 0.000 description 2
- 229910052710 silicon Inorganic materials 0.000 description 2
- 239000010703 silicon Substances 0.000 description 2
- 238000004611 spectroscopical analysis Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 208000001613 Gambling Diseases 0.000 description 1
- 239000012491 analyte Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 239000011248 coating agent Substances 0.000 description 1
- 238000000576 coating method Methods 0.000 description 1
- 229910052681 coesite Inorganic materials 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 229910052906 cristobalite Inorganic materials 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000011810 insulating material Substances 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035699 permeability Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 239000000377 silicon dioxide Substances 0.000 description 1
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 229910052682 stishovite Inorganic materials 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
- 229910052905 tridymite Inorganic materials 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D5/00—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable
- G01D5/26—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable characterised by optical transfer means, i.e. using infrared, visible, or ultraviolet light
- G01D5/32—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable characterised by optical transfer means, i.e. using infrared, visible, or ultraviolet light with attenuation or whole or partial obturation of beams of light
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
A supervised learning-based self-interference micro-ring resonant cavity sensing classification identification method comprises the following steps: (1) collecting training data; (2) training a BP neural network sensing data detection model; (3) collecting test data; (4) BP neural network sensing data detection model test: and (3) inputting the test data obtained in the step (3), namely the transmission extinction value within a certain wavelength range, into the trained neural network, and outputting two label values of the target to be tested. The invention can realize the formation of N-type basic substances by the output of the neural networkNAnd (4) identifying and classifying 1 different combined components to obtain an identification and classification result of the target substance to be detected.
Description
Technical Field
The invention relates to a self-interference micro-ring resonant cavity sensing classification identification method based on supervised learning, and belongs to the field of optical microcavity sensing.
Background
The whispering gallery mode optical microcavity has ultrahigh Q factor and extremely small mode volume, and is widely applied to the fields of active optical devices, optical signal processing, optical interconnection, low-energy nonlinear optics, optical and substance interaction, sensing and the like. The echo wall optical microcavity sensor can greatly enhance the interaction between an optical field and a substance, effectively improve the detection sensitivity, and has wide application prospect in the sensing fields of single nano-particles, biomolecules and the like. In the echo wall optical microcavity, an optical field is limited in the resonance microcavity, but part of energy is still leaked into the environment through the evanescent field, and the substance to be detected interacts with the evanescent field, so that the change of an optical mode, including resonance mode movement, mode line width widening, degenerate mode splitting and the like, can be caused. These arise from the responsive interaction of the mode field to changes in the probe material, collectively referred to as the dispersion sensing mechanism. When the nano-particles are absorbable, the energy of resonance is lost due to edge scattering, and the line width of the whispering gallery microcavity mode is widened. This dissipative interaction creates a sensing mechanism known as a dissipative sensing mechanism.
When the traditional echo wall microcavity is used as a sensor, the frequency spectrum shift is almost consistent at each resonant wavelength mode, so that for the traditional echo wall microcavity sensing, no matter a dispersion sensing mechanism or a dissipation sensing mechanism, only a certain mode is selected to realize sensing measurement, and relevant sensing information at other multiple wavelength modes is not needed. However, such single-mode sensing measurements clearly do not classify and identify different target substances. Such as identification and classification of biological components, in more sophisticated measurements of biological components, not only specific measurements but also accurate identification and classification are required. Based on the echo wall microcavity resonance frequency shift and the echo wall mode number in the free spectral region of these biological components, in combination with the neural network, classification of biological components is achieved with an average accuracy of 97.3% (document 1: e.a.tcherniavskaia, v.a.saetchnikov, Application of neural networks for classification of biological compounds from the characteristics of wireless-gambling-mode optical resources, journal Applied Spectroscopy,2011, 78, pp.457-460, i.e., e.a.tcherniavskaia, v.a.saetchnikov, based on the classification of wall microcavity resonance biological components of the neural network, Application of optical Spectroscopy, 78,457 (2011)). But the classification precision is obviously not general and is difficult to be widely adopted in practice.
In conclusion, in the sensing classification identification based on the echo wall microcavity, a single-mode detection mode is still adopted, so that the detection accuracy is not high and the generality is not high. Therefore, in order to realize a general high classification accuracy method, a new multimode sensing classification identification method must be explored.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention aims to realize a self-interference micro-ring resonant cavity sensing classification and identification method based on supervised learning, and the classification precision is higher.
In order to solve the technical problems, the invention provides the following technical scheme:
a supervised learning-based self-interference micro-ring resonant cavity sensing classification identification method comprises the following steps:
(1) training data acquisition, the process is as follows:
1-1, firstly, adopting a plurality of groups of training data for training an artificial neural network sensing data detection model, wherein each group of training data consists of a transmission extinction value in a certain wavelength range and two pre-known training label values corresponding to the transmission extinction value, and respectively serving as an input value and a target output value of a BP neural network to train the BP neural network so as to establish a mapping relation between the transmission extinction value and the target output value of the BP neural network; by changing a target substance to be measured and collecting a transmission extinction value in a corresponding certain wavelength range, a plurality of groups of training data can be obtained;
1-2, after collecting enough groups of training data, carrying out normalization processing on the collected transmission extinction value in a certain wavelength range and two training label values which are known in advance and correspond to the transmission extinction value, and taking a processed data set as a final training data set;
(2) the BP neural network sensing data detection model training comprises the following steps:
2-1, using the transmission extinction value within a certain wavelength range processed in the step 1 as input data, and using the two training label values processed in the step 1 as output data;
2-2 training a BP neural network sensing data detection model, and establishing and storing a mapping relation between a transmission extinction value in a certain wavelength range and two corresponding training label values.
(3) Test data acquisition, the process is as follows:
3-1, placing the whole detection system in a measurement environment, detecting the label values of two target substances to be detected under three different conditions, and collecting transmission extinction values within a certain wavelength range by using a photoelectric detector and an oscilloscope when the detection system is excited by using an adjustable laser light source;
3-2, carrying out normalization processing on the acquired transmission extinction value, and taking the transmission extinction value as a test data set;
(4) BP neural network sensing data detection model test: and (3) inputting the test data obtained in the step (3), namely the transmission extinction value within a certain wavelength range, into the trained neural network, and outputting two label values of the target to be tested.
Further, the self-interference micro-ring resonant cavity comprises an input waveguide, a micro-ring resonant cavity, an output waveguide and an optical detection arm waveguide, wherein the input waveguide and the output waveguide are respectively coupled with the micro-ring resonant cavity and are arranged on two sides of the micro-ring resonant cavity, one end of the input waveguide is a light source access end of the whole optical sensor, the other end of the input waveguide is connected with an input end of the optical detection arm waveguide at the coupling position of the input waveguide and the micro-ring resonant cavity, the output end of the optical detection arm waveguide is connected with one end of the output waveguide at the coupling position of the output waveguide and the micro-ring resonant cavity, and the other end of the output waveguide is a sensing signal exit end.
Furthermore, two materials sensitive to the detected target substance are respectively coated on the upper surfaces of the two sections of micro-ring waveguides or the two sections of optical detection arm waveguides, the output light of the tunable laser is incident from one end of the input waveguide and coupled with the micro-ring resonant cavity, one portion is coupled into the micro-ring resonator and the other portion exits the other end of the input waveguide and passes through the optical probe arm into the output waveguide, this portion of light, due to the coupling between the output waveguide and the microresonator, is coupled into the microresonator again, and a part of light in the part interferes with a part of light coupled out from the micro-ring resonant cavity and then is emitted out from the other end of the output waveguide, because the refractive indexes of the sensitive materials of the two corresponding target substances to be detected are changed, thereby causing the refractive index of the two sections of micro-ring waveguides or the two sections of detection arm waveguides to be changed, and further causing the change of the emergent frequency spectrum; in a typical transmission spectrum of a self-interference micro-ring resonant cavity, due to different interaction between each target substance sensitive material to be detected and the target substance to be detected, the refractive index of a micro-ring waveguide or a detection arm waveguide is changed differently at each wavelength, and further, the transmission extinction at different resonance wavelengths is changed differently; therefore, for the self-interference micro-ring resonant cavity optical sensor, the artificial neural network sensing data detection model is established to realize identification and classification by extracting the change of effective sensing information (transmission extinction at a plurality of resonant wavelengths) in a certain wavelength range.
Further, the artificial neural network is supervised learning, needs to be trained before classification and identification, and if two substances and their combination need to be identified and classified, the training data involved in the training process is divided into three cases: (1) when both substances are present, i.e. the combination of them, the training labels are respectively marked as 1 and 1; (2) when only the first substance is present, the training labels are respectively marked as 1 and 0; (3) when only the second substance is present, the training labels are noted as 0 and 1, respectively. In each case, effective sensing information (transmission extinction values at different resonance wavelengths) of a known training label is input for training, and a trained neural network is stored after training in all three cases; and then testing effective sensing information (transmission extinction values at different resonance wavelengths) of the test data, and finally obtaining the identification classification result of the target substance to be tested through neural network output (label value).
The invention has the beneficial effects that: the invention relates to a supervised learning-based self-interference micro-ring resonant cavity sensing classification identification method, which comprises the steps of respectively coating two corresponding detection target substance sensitive materials on the upper surfaces of N sections of annular waveguides or two sections of optical detection arm waveguides, extracting effective sensing information (transmission extinction at a plurality of resonance wavelengths) in a certain wavelength range by detecting the emergent frequency spectrum of the detection target substance sensitive materials, and establishing an artificial neural network sensing data detection model to realize identification classification; the used artificial neural network is supervised learning, needs to be trained before classification and identification, then tests effective sensing information (transmission extinction values at different resonance wavelengths) of test data, and finally outputs (label values) through the neural network to realize 2 formed by N types of basic substancesNAnd (4) identifying and classifying 1 different combined components to obtain an identification and classification result of the target substance to be detected.
Drawings
FIG. 1 is a schematic diagram of a self-interference micro-ring resonant cavity sensor;
FIG. 2 is a typical emission spectrum of the self-interference micro-ring resonator sensor in the wavelength range 1400-1600 nm;
FIG. 3 is a typical exit spectrum from the wavelength range 1520-1540nm of an interference-type micro-ring resonator sensor;
FIG. 4 is a sensing classification recognition model based on BP neural network;
FIG. 5 contains test results of a plurality of test data of three components via a neural network.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the embodiments of the present invention are not limited thereto.
Referring to fig. 1 to 5, a supervised learning-based self-interference micro-ring resonant cavity sensing classification identification method includes the following steps:
(1) training data acquisition, the process is as follows:
1-1, firstly, adopting a plurality of groups of training data for training an artificial neural network sensing data detection model, wherein each group of training data consists of a transmission extinction value in a certain wavelength range and two pre-known training label values corresponding to the transmission extinction value, and respectively serving as an input value and a target output value of a BP neural network to train the BP neural network so as to establish a mapping relation between the transmission extinction value and the target output value of the BP neural network; similarly, by changing the target substance to be measured (such as the concentration of gas) and collecting the transmission extinction value within a corresponding certain wavelength range, a plurality of groups of training data can be obtained;
1-2, after collecting enough groups of training data, carrying out normalization processing on the collected transmission extinction value in a certain wavelength range and two training label values which are known in advance and correspond to the transmission extinction value, and taking a processed data set as a final training data set;
(2) the BP neural network sensing data detection model training comprises the following steps:
2-1, using the transmission extinction value within a certain wavelength range processed in the step 1 as input data, and using the two training label values processed in the step 1 as output data;
2-2 training a BP neural network sensing classification recognition model, and establishing and storing a mapping relation between a transmission extinction value in a certain wavelength range and two corresponding training label values.
(3) Test data acquisition, the process is as follows:
3-1, placing the whole detection system in a measurement environment, detecting the label values of two target substances to be detected under three different conditions, and collecting transmission extinction values within a certain wavelength range by using a photoelectric detector and an oscilloscope when the detection system is excited by using an adjustable laser light source;
3-2, carrying out normalization processing on the acquired transmission extinction value, and taking the transmission extinction value as a test data set;
(4) BP neural network sensing data detection model test: inputting the test data obtained in the step 3, namely the transmission extinction value within a certain wavelength range, into the trained neural network, and outputting two label values of the target to be tested;
in the step (1), feasibility of the supervised learning based sensing classification identification method is proved from the perspective of theoretical simulation in the embodiment of the present invention. When the target to be measured is three different components, useful sensing information (transmission extinction value) on a corresponding transmission spectrum is obtained from theoretical simulation and used as training data, after a neural network is trained, the useful sensing information (transmission extinction value) to be measured outside a training data set is tested, tag values of two gas concentrations to be measured are output and compared with an initial preset result.
The assumed models of the effective refractive index of the waveguide as a function of wavelength are of the lorentzian type and of the gaussian type, respectively. With two gas concentrations C1And C2Firstly, theoretical simulation is carried out to obtain a corresponding frequency spectrum, and useful sensing information, namely a transmission extinction value in a certain frequency spectrum, is extracted. As shown in FIG. 2, the emission spectrum of two gas components in three combinations is shown in the wavelength range from 1400-1600 nm. FIG. 3 further shows the exit spectrum of FIG. 2 over the wavelength range from 1520nm to 1540 nm. It can be seen that the transmission extinction at different wavelength modes varies quite differently for the three different gas components.
Fig. 4 shows a sensory classification recognition model based on a BP neural network. It is divided into an input layer, a hidden layer and an output layer. The second layer is a hidden layer, and the node number value of the hidden layer is too low, so that the condition of non-convergence in the learning process can occur. On the contrary, if the node number of the hidden layer is too high, the network error can be reduced, the precision is improved, but the network is also complicated, so that the training time of the network is prolonged, and the tendency of overfitting is caused. The embodiment of the invention constructs a sensing data measurement model based on a BP neural network, and the number of hidden layer nodes is obtained by adopting an empirical formula. The third layer is an output layer. The node number of the output layer is 2, and the two output results are two target gas label values (L)1,L2). Training the neural network by the processed training data, specifically, taking the normalized transmission extinction value in the training data as input and the corresponding known two target gas label values as two outputs, so that the neural network learns and protectsAnd storing a mapping rule between the two.
In the step (1), training is carried out according to three components: (1) at two gas concentrations C1And C2Is taken to be [0.000500869:0.00001999:0.000980629]Its initial value is 0.000500869, maximum value is 0.000980629, value interval is 0.00001999, label value (L)1,L2) Number of training sets 625 to obtain a complete training set with gas concentration C ═ 1,11Gas concentration C from 0.000500869 to 0.0009806292Temporarily not changing until the first gas concentration completes a change cycle; (2) is a gas concentration C1The value is [0.000501376:0.00000099:0.000995386]Its initial value is 0.000501376, maximum value is 0.000995386, and value interval is 0.00000099, at this time, gas concentration C is2Is 0, the tag value (L)1,L2) The number of training groups is 500 groups (1, 0); (3) is a gas concentration of C1Gas concentration C ═ 02Is taken to be [0.000501376:0.00000099:0.000995386]Its initial value is 0.000501376, maximum value is 0.000995386, value interval is 0.00000099, (L)1,L2) The number of training sets is 500 sets (0, 1). These training data were obtained by taking two gas concentrations C1And C2The preset value of (2) is substituted into the theoretical expression of the transmission coefficient to obtain a corresponding transmission spectrum, so as to obtain a corresponding transmission extinction value (in all simulation spectrums, a transmission extinction threshold value of 0.90 is selected, that is, only the transmission extinction value lower than the threshold value is used as effective sensing data). And carrying out normalization processing on the transmission extinction value to be used as the input of the neural network, and taking the corresponding two gas label values as output prediction values to train the neural network.
In the steps (3) and (4), the transmission extinction values corresponding to the three combinations of the target substances to be tested are used for testing, as shown in fig. 5. (1) Gas concentration C1And C2The value is [0.0005:0.00002:0.0009 ]]The initial value is 0.0005, the maximum value is 0.0009, the value interval is 0.00002, and at this time, two output label values (L) tested by the neural network1,L2) Two gases in this case are indicated (1,1)Are both present; (2) gas concentration C1The value is [0.0005:0.0001:0.001 ]]Initial value of 0.0005, maximum value of 0.001, interval of 0.0001, gas concentration C2The value is 0, and two output label values (L) tested by the neural network at the moment1,L2) (1,0), indicating that in this case the 1 st gas is present; (3) gas concentration C2Value of 0, gas concentration C2The value is [0.0005:0.0001:0.001 ]]The initial value is 0.0005, the maximum value is 0.001, the value interval is 0.0001, and at this time, two output label values (L) tested by the neural network are obtained1,L2) Which indicates that the 2 nd gas is present in this case (0, 1). The above test data are obtained in particular by the two gas concentrations C1And C2And substituting the transmission spectrum into the theoretical model to obtain a corresponding transmission extinction value. And carrying out normalization processing on the transmission extinction values to respectively obtain corresponding test data. The simulation result proves that the test result in the embodiment of the invention can effectively identify and classify the three different components, and has excellent prediction performance.
When the basic components of the target substance to be detected are 3 or more than 3, the corresponding label value is only required to be increased. For example, when the basic components of the target substance to be detected are 3, the total component types are 7, and therefore the corresponding output tag value is 7 values of 001-.
In this embodiment, table 1 shows parameters of a self-interference micro-ring resonator based on a Silicon On Insulator (SOI) material;
TABLE 1
In the embodiment of the present invention, the self-interference micro-ring resonator structure is shown in fig. 1, and is different from a common two-parallel waveguide coupling micro-ring resonator structure, and the structure connects an upper coupling region and a lower coupling region in the two-parallel waveguide coupling micro-ring structure by using an additional probe waveguide arm. The output is essentially the result of interference of the inner micro-ring resonator and the feedback loop micro-ring resonator. The detection substance causes the micro-phase change of the micro-ring waveguide or the detection arm waveguide, and the external coupling coefficient of the micro-ring and the whole U-shaped detection arm waveguide is changed rapidly, so that the emergent frequency spectrum of the micro-ring waveguide or the detection arm waveguide is changed.
The designed self-interference micro-ring resonant cavity sensor chip is composed of a silicon-based (SOI) waveguide based on an insulating material, and a substrate material is SiO2The parameters are set as shown in Table 1, and the effective refractive index n of the waveguideeff2.45, radius of micro-ring R30 μm, initial length L of waveguide of U-shaped probe armW250 mu m, the directional coupling energy coupling coefficient k is 0.5, and the waveguide loss coefficient α is 0.1 dB/cm.
The designed self-coherent micro-ring resonant cavity sensing chip structure is shown in figure 1. For the self-interference micro-ring resonant cavity structure, two sensitive materials are respectively coated on the surfaces of two sections of micro-ring waveguides, and the two sensitive materials respectively have relatively high detection sensitivity on two sensing detection targets. The sensor exposes only the microring to the probe target without any contact of the probe arm waveguide with the probe target. Therefore, according to the dissipative sensing principle of the self-coherent micro-ring resonant cavity, the typical transmission spectrum of the emergent spectrum can still be obtained. Two gas sensitive materials are respectively placed on the surfaces of the left and right sections of the micro-ring waveguides. Due to the interaction of gas molecules and corresponding sensitive materials, the change delta n of the refractive index of the cladding of the micro-ring waveguide is causedc(λ, C), which is related to the wavelength λ and the gas concentration C. When the concentration of a gas to be measured is C, the effective refractive index change delta n of a certain wavelength lambda modeeff(lambda, C) is represented by,
Δneff(λ,C)＝SwaveguideΔnc(λ,C)
wherein SwaveguideFor dielectric waveguide sensitivity, Δ nc(λ, C) is the change in refractive index of the cladding of the gas-sensitive material at a certain wavelength λ mode at a certain gas concentration C. Dielectric waveguide sensitivity is related to waveguide structure and materials, and also to cladding index changes,
if the gas can be absorbed by the gas-sensitive material,. DELTA.nc(lambda, C) can be represented as,
Δnc(λ,C)＝F(λ,λ0)(λ0)νtSbpC
wherein λ0Is the central wavelength of the absorption band, F (lambda )0) Is the proportionality coefficient between absorption and refractive index change, vtIs the total number of gas-sensitive material molecules per unit volume of the polymer, SbIs the reagent and analyte binding constant and p is the polymer permeability constant.
As shown in fig. 1, two sections of micro-ring waveguides are coated with different gas sensitive materials, and the interaction between two gases and their corresponding sensitive materials is completely different, so that the two targets to be measured (the concentrations of the two gases) have completely different influences on the effective refractive index of the waveguide at the wavelength of each mode. It is assumed that the two waveguide effective index contributions have two different analytical models, one gaussian, one lorentz,
where C is1And C2Representing the concentrations of two target gases to be detected, the parameter values in the two formulas are carefully set so that the refractive index change magnitude of the two formulas is consistent with the waveguide refractive index change magnitude in actual gas detection.
The change of the refractive index of the waveguide is substituted into the transmission coefficient expression of the self-interference micro-ring resonant cavity,
for simplicity, the waveguide and microring are assumed to have the same propagation constant βW＝βR＝(2π/λ)neff-i α, where the initial probe arm waveguide index nL＝nR＝neffAnd α is the waveguide loss coefficient, here βR1＝(2π/λ)(neff+Δneff1)-iα，βR2＝(2π/λ)(neff+Δneff2) I α represents the propagation constant of the waveguide after the interaction of two sections of waveguide coated with two gas sensitive materials and two gas molecules respectively1And C2Firstly, theoretical simulation is carried out to obtain a corresponding frequency spectrum, and useful sensing information, namely a transmission extinction value in a certain frequency spectrum, is extracted. As shown in FIG. 2, the emission spectrum of two gas components in three combinations is shown in the wavelength range from 1400-1600 nm. FIG. 3 further shows the exit spectrum of FIG. 2 over the wavelength range from 1520nm to 1540 nm. It can be seen that the transmission extinction at different wavelength modes varies quite differently for the three different gas components. On the basis, a neural network sensing classification recognition model is constructed, sensing data processing is carried out, and classified sensing detection targets are recognized, as shown in fig. 4. And taking the extracted transmission extinction value as the input of a neural network, taking the two target gas prediction tag values as expected outputs, and training the neural network. And randomly finding a test set corresponding to the same label value outside the training set, namely corresponding to the useful sensing information of the two label values, and testing by using the trained neural network. The test result is shown in fig. 5, the theoretical original data is well consistent with the detection result obtained through the neural network, and the feasibility of the detection, identification and classification method is illustrated.
In the embodiment of the present invention, the parameters of the neural network are shown in table 2:
TABLE 2
The above-mentioned embodiments are only used to help understanding the method and its core idea of the present invention, and not to limit the same, and any other changes, modifications, substitutions, combinations, and simplifications that do not depart from the spirit and principle of the present invention should be regarded as equivalent substitutions and shall be included within the scope of the present invention.
Claims (4)
1. A supervised learning-based self-interference micro-ring resonant cavity sensing classification identification method is characterized by comprising the following steps:
(1) training data acquisition, the process is as follows:
1-1, firstly, adopting a plurality of groups of training data for training an artificial neural network sensing data detection model, wherein each group of training data consists of a transmission extinction value in a certain wavelength range and two pre-known training label values corresponding to the transmission extinction value, and respectively serving as an input value and a target output value of a BP neural network to train the BP neural network so as to establish a mapping relation between the transmission extinction value and the target output value of the BP neural network; by changing a target substance to be measured and collecting a transmission extinction value in a corresponding certain wavelength range, a plurality of groups of training data can be obtained;
1-2, after collecting enough groups of training data, carrying out normalization processing on the collected transmission extinction value in a certain wavelength range and two training label values which are known in advance and correspond to the transmission extinction value, and taking a processed data set as a final training data set;
(2) the BP neural network sensing data detection model training comprises the following steps:
2-1, using the transmission extinction value within a certain wavelength range processed in the step 1 as input data, and using the two training label values processed in the step 1 as output data;
2-2 training a BP neural network sensing data detection model, and establishing and storing a mapping relation between a transmission extinction value in a certain wavelength range and two corresponding training label values;
(3) test data acquisition, the process is as follows:
3-1, placing the whole detection system in a measurement environment, detecting the label values of two target substances to be detected under three different conditions, and collecting transmission extinction values within a certain wavelength range by using a photoelectric detector and an oscilloscope when the detection system is excited by using an adjustable laser light source;
3-2, carrying out normalization processing on the acquired transmission extinction value, and taking the transmission extinction value as a test data set;
(4) BP neural network sensing data detection model test: and (3) inputting the test data obtained in the step (3), namely the transmission extinction value within a certain wavelength range, into the trained neural network, and outputting two label values of the target to be tested.
2. The supervised learning based self-interference micro-ring resonator sensing classification and identification method of claim 1, wherein the self-interference micro-ring resonator comprises an input waveguide, a micro-ring resonator, an output waveguide and an optical detection arm waveguide, the input waveguide and the output waveguide are respectively coupled with the micro-ring resonator and are disposed at two sides of the micro-ring resonator, one end of the input waveguide is a light source access end of the whole optical sensor, the other end of the input waveguide is connected with an input end of the optical detection arm waveguide at a coupling position of the input waveguide and the micro-ring resonator, the output end of the optical detection arm waveguide is connected with one end of the output waveguide at a coupling position of the output waveguide and the micro-ring resonator, and the other end of the output waveguide is a sensing signal exit end.
3. The supervised learning-based self-interference type micro-ring resonant cavity sensing classification and identification method as recited in claim 1 or 2, wherein two materials sensitive to the target substance to be detected are respectively coated on the upper surfaces of two sections of micro-ring waveguides or two sections of optical detection arm waveguides, output light of the tunable laser enters from one end of the input waveguide and is coupled with the micro-ring resonant cavity, then a part of the output light is coupled into the micro-ring resonant cavity, and another part of the output light exits from the other end of the input waveguide and enters the output waveguide through the optical detection arm, the part of the light is coupled into the micro-resonant cavity again due to the coupling effect between the output waveguide and the micro-resonant cavity, and then exits from the other end of the output waveguide after the part of the light in the part of the light interferes with a part of the light coupled out of the micro-ring resonant cavity, and the refractive indexes of the two materials sensitive to-object substances to, thereby causing the refractive index of the two sections of micro-ring waveguides or the two sections of detection arm waveguides to be changed, and further causing the change of the emergent frequency spectrum; in a typical transmission spectrum of a self-interference micro-ring resonant cavity, due to different interaction between each target substance sensitive material to be detected and the target substance to be detected, the refractive index of a micro-ring waveguide or a detection arm waveguide is changed differently at each wavelength, and further, the transmission extinction at different resonance wavelengths is changed differently; therefore, for the self-interference micro-ring resonant cavity optical sensor, the change of effective sensing information is extracted within a certain wavelength range, and an artificial neural network sensing data detection model is established to realize identification and classification.
4. The supervised learning-based self-interference micro-ring resonator sensing classification and identification method as recited in claim 1 or 2, wherein the artificial neural network is supervised learning, and needs to be trained before classification and identification, and if two substances and their combination need to be identified and classified, the training data involved in the training process is divided into three cases: (1) when both substances are present, i.e. the combination of them, the training labels are respectively marked as 1 and 1; (2) when only the first substance is present, the training labels are respectively marked as 1 and 0; (3) when only the second substance is present, the training labels are respectively noted as 0 and 1; in each case, effective sensing information input of a known training label is used for training, and the trained neural network is stored after the training in all three cases; and then testing the effective sensing information of the test data, and finally outputting through a neural network to obtain the identification classification result of the target substance to be tested.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010156324.XA CN111721336A (en) | 2020-03-09 | 2020-03-09 | Self-interference micro-ring resonant cavity sensing classification identification method based on supervised learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010156324.XA CN111721336A (en) | 2020-03-09 | 2020-03-09 | Self-interference micro-ring resonant cavity sensing classification identification method based on supervised learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111721336A true CN111721336A (en) | 2020-09-29 |
Family
ID=72564070
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010156324.XA Pending CN111721336A (en) | 2020-03-09 | 2020-03-09 | Self-interference micro-ring resonant cavity sensing classification identification method based on supervised learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111721336A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113916807A (en) * | 2021-08-27 | 2022-01-11 | 浙江工业大学 | Micro-cavity optical frequency comb gas concentration sensing measurement method |
| CN114199785A (en) * | 2021-11-18 | 2022-03-18 | 国网浙江省电力有限公司诸暨市供电公司 | Echo wall micro-cavity sensing method based on GAN data enhancement |
| CN118192010A (en) * | 2024-02-01 | 2024-06-14 | 重庆邮电大学 | Silicon-based light-operated micro-ring resonator for in-situ training of optical neural network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101183899A (en) * | 2007-12-19 | 2008-05-21 | 天津大学 | BP network based pipeline security identifying method for optical fiber pipeline leakage monitoring device |
| CN105675545A (en) * | 2016-01-15 | 2016-06-15 | 浙江工业大学 | High-sensitivity intensity detection method based on self-interference type micro resonator cavity light sensor |
| CN110554006A (en) * | 2019-09-04 | 2019-12-10 | 中国科学技术大学 | Multi-mode measurement method based on self-interference micro-ring resonant cavity sensor |
| US20200019851A1 (en) * | 2018-07-10 | 2020-01-16 | The George Washington University | Optical convolutional neural network accelerator |
-
2020
- 2020-03-09 CN CN202010156324.XA patent/CN111721336A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101183899A (en) * | 2007-12-19 | 2008-05-21 | 天津大学 | BP network based pipeline security identifying method for optical fiber pipeline leakage monitoring device |
| CN105675545A (en) * | 2016-01-15 | 2016-06-15 | 浙江工业大学 | High-sensitivity intensity detection method based on self-interference type micro resonator cavity light sensor |
| US20200019851A1 (en) * | 2018-07-10 | 2020-01-16 | The George Washington University | Optical convolutional neural network accelerator |
| CN110554006A (en) * | 2019-09-04 | 2019-12-10 | 中国科学技术大学 | Multi-mode measurement method based on self-interference micro-ring resonant cavity sensor |
Non-Patent Citations (1)
| Title |
|---|
| DONG HU 等: "Multi-Parameter Sensing in a Multimode", 《SENSORS》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113916807A (en) * | 2021-08-27 | 2022-01-11 | 浙江工业大学 | Micro-cavity optical frequency comb gas concentration sensing measurement method |
| CN113916807B (en) * | 2021-08-27 | 2023-07-14 | 浙江工业大学 | Micro-cavity optical frequency comb gas concentration sensing measurement method |
| CN114199785A (en) * | 2021-11-18 | 2022-03-18 | 国网浙江省电力有限公司诸暨市供电公司 | Echo wall micro-cavity sensing method based on GAN data enhancement |
| CN114199785B (en) * | 2021-11-18 | 2023-09-26 | 国网浙江省电力有限公司诸暨市供电公司 | Echo wall microcavity sensing method based on GAN data enhancement |
| CN118192010A (en) * | 2024-02-01 | 2024-06-14 | 重庆邮电大学 | Silicon-based light-operated micro-ring resonator for in-situ training of optical neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Myrick et al. | Comparison of some fiber optic configurations for measurement of luminescence and Raman scattering | |
| Patil et al. | Comprehensive and analytical review on optical fiber refractive index sensor | |
| CN111721336A (en) | Self-interference micro-ring resonant cavity sensing classification identification method based on supervised learning | |
| US20160131580A1 (en) | Optical analyte sensor | |
| US11619580B2 (en) | Resonator-based ion-selective sensor | |
| Zhang et al. | Fiber Loop Ring-Down Refractive Index Sensor Based on High-$ Q $ Photonic Crystal Cavity | |
| Li et al. | Detection limit analysis of optical fiber sensors based on interferometers with the Vernier-effect | |
| CN109253986B (en) | Double-ring optical sensor of cascade Fourier transform spectrometer | |
| Gounaridis et al. | High performance refractive index sensor based on low Q-factor ring resonators and FFT processing of wavelength scanning data | |
| WO2013053930A1 (en) | Resonant biosensor | |
| CN111523376B (en) | Self-interference micro-ring resonant cavity sensing classification identification method based on unsupervised learning | |
| Iftekharul Ferdous et al. | Unlocking insights of oil derivatives with terahertz spectrum analysis: the hybrid refractive index rectangular core photonic crystal fiber perspective sensing | |
| Petermann et al. | All-polymer whispering gallery mode sensor system | |
| Mao et al. | Reflective temperature sensor based on a fiber Bragg grating combined with a fiber loop ringdown technique | |
| Dong et al. | Displacement sensing in a multimode SNAP microcavity by an artificial neural network | |
| Zhu et al. | Vernier effect-based optical fiber sensor for dynamic sensing using a coarsely resolved spectrometer | |
| Gill et al. | On-chip absorption spectroscopy enabled by graded index fiber tips | |
| Dong et al. | Rapid and high-precision displacement sensing based on the multiple mode dip areas in a SNAP microresonator | |
| Mallik et al. | Study of the influence of the agarose hydrogel layer thickness on sensitivity of the coated silica microsphere resonator to humidity | |
| CN114184296A (en) | Multimode micro-ring resonant cavity temperature sensing processing method and device based on transfer learning | |
| Alameri et al. | Design in line lma-10 photonic crystal fiber mzi for detection and investigation of beta thalassemia | |
| Arifin et al. | Low‐Cost High‐Sensitive Plastic Optical Fiber‐Based Sensor for Detection of CO (NH2) 2 Urea | |
| Vishwaraj et al. | Machine learning assisted strip waveguide Bragg gratings design for refractive index-based biosensing applications | |
| TWI467158B (en) | Application method of optical filter spectral linewidth | |
| Thompson | Cavity‐Enhanced Spectroscopy in Condensed Phases: Recent Literature and Remaining Challenges |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200929 |