KR20230021909A - System and method for anomaly detection of crane equipment using artificial intelligence - Google Patents

Abstract

Disclosed are a system and method for detecting abnormality of crane equipment using artificial intelligence (AI). According to the present invention, as a system for detecting abnormalities in crane equipment that generates vibration signals according to operation, a system for detecting abnormality of crane equipment using AI may comprise: a sensor for detecting the vibration signal; and a computing device which learns pre-input normal vibration data and abnormal vibration data, and detects an abnormal state of the crane equipment based on the learned result value and the vibration signal.

Description

Anomaly detection system and method of crane equipment using artificial intelligence {SYSTEM AND METHOD FOR ANOMALY DETECTION OF CRANE EQUIPMENT USING ARTIFICIAL INTELLIGENCE}

The present invention relates to a system and method for detecting abnormalities in crane equipment using artificial intelligence. More specifically, it relates to a system and method capable of learning vibrations occurring in crane equipment and detecting abnormality of crane equipment based on this.

Conventional methods for detecting abnormalities in heavy equipment have been built through a method of processing vibration signals based on expert experience and engineering knowledge and setting a specific threshold for vibration signals collected based on business rules. However, there are many types of factors that indicate the state of facilities, and there are many facilities managed by facility managers, so it is difficult to set an appropriate threshold value considering the characteristics of facilities. Since specific background knowledge is essential, it is difficult for non-experts to determine what kind of defect has occurred.

In addition, the conventional alarm method through the threshold value is difficult to distinguish between a signal that temporarily exceeds the threshold value due to an external shock and a signal caused by a defect, and a warning sounds due to an external shock even though there is no actual defect in the equipment, resulting in damage to the facility. It could cause economic loss such as downtime and maintenance.

Therefore, it is difficult to provide accurate monitoring performance only by determining whether an alarm associated with a vibration signal exceeds a specific threshold, and it is important to propose an accurate countermeasure to overcome a failure of heavy equipment. Currently, a vibration monitoring system has been partially implemented in Korea, but the development of a system and method capable of learning vibration signals and detecting abnormal conditions according to them as well as classifying them and suggesting appropriate countermeasures is insufficient. .

KR102138279B1 (Title of invention: Apparatus and method for monitoring vibration condition of rotating equipment using deep learning-based time series analysis)

An object of the present invention is to automatically detect abnormal conditions of crane equipment by detecting vibrations generated in crane equipment and learning them.

In addition, an object of the present invention is to automatically classify a vibration signal when an abnormal state is detected and guide a user to a corresponding countermeasure.

In order to solve the above problems, the present invention is a system for detecting abnormality of crane equipment that generates a vibration signal according to the operation, a sensor for detecting the vibration signal and pre-input normal vibration data and abnormal vibration data It may include a computing device that learns and detects an abnormal state of the crane equipment based on the learned result value and the vibration signal.

In addition, the computing device may classify the abnormal state into a plurality of modes, and the plurality of modes may be determined in advance based on whether coping methods according to the abnormal state are the same.

In addition, the computing device applies a Fast Fourier Transform (FFT) to the steady vibration data and the abnormal vibration data, derives a first difference value to the steady vibration data and the abnormal vibration data to which the FFT is applied, and the first difference value can learn

In addition, the computing device extracts a first vibration signal for a first period and a second vibration signal for a second period with respect to the vibration signal, and the computing device extracts the first vibration signal and the second vibration signal. FFT is applied to, and a second difference value between the first vibration signal and the second vibration signal to which the FFT is applied is derived, and the ideal state of the crane equipment is based on the similarity between the second difference value and the first difference value. can detect

In addition, when the computing device detects an abnormal state of the crane equipment, performs a K-center clustering process on the second difference value, and finally generates two clusters based on the K-center clustering process, , the similarity can be derived based on the two clusters.

In addition, in order to solve the above problems, the present invention is a step of learning the normal vibration data generated from the crane equipment in a normal state and the abnormal vibration data generated from the crane equipment in an abnormal state, the vibration signal generated from the target crane equipment Receiving and detecting an abnormal state of the target crane equipment based on the learned result value and the vibration signal, wherein the learning step is performed using FFT (Fast Fourier Transform) on the normal vibration data and the abnormal vibration data. , deriving a first difference value to the steady vibration data and abnormal vibration data to which the FFT is applied, and learning the first difference value.

In addition, in order to solve the above problems, the present invention is a non-transitory computer readable medium storing instructions, wherein the instructions, when executed by a processor, cause the processor to perform the above-described method, a non-transitory computer readable medium It can be a possible medium.

The present invention learns the difference of vibration data to which FFT (Fast Fourier Transform) is applied, and has an effect of enabling faster learning with less resources.

In addition, the present invention detects the abnormal state of the crane equipment based on the difference between the normal vibration data and the abnormal vibration data, so that the abnormal state can be diagnosed more quickly with less resources.

In addition, the present invention detects abnormal conditions that may occur in crane equipment through artificial intelligence, thereby automatically and immediately detects the failure of the crane, thereby promoting user safety and preventing large-scale accidents.

In addition, the present invention automatically classifies the type of abnormal condition and guides the user to an appropriate countermeasure accordingly, so that an emergency response can be quickly responded and a large-scale accident can be prevented in advance.

In addition, the present invention has an effect of generating accurate vibration data by protecting the sensor from external contaminants or salt by proposing a sensor including a unique coating layer.

In addition, the present invention has an effect of enabling more accurate detection and classification of an abnormal state by proposing an abnormal state detection and classification algorithm using a specific correction constant.

The effects obtainable according to the present invention are not limited to the effects mentioned above, and other effects not mentioned are clearly understood by those skilled in the art from the outer surface of the protection unit below. It could be.

1 shows an anomaly detection system for crane equipment using artificial intelligence according to the present invention.
2 illustrates a computing device in accordance with the present invention.
3 shows a functional configuration included in a memory according to the present invention.
4 shows a graph to which the K-center clustering process performance module according to the present invention is applied.
5 is an example of comparing Euclidean distance values for each mode according to the present invention.
6 to 9 schematically show an abnormal state detection process according to the present invention.
10 schematically illustrates the configuration of a sensor according to the present invention.
11 and 12 show a method for detecting abnormalities in crane equipment using artificial intelligence according to the present invention.
The accompanying drawings, which are included as part of the detailed description to aid understanding of the present specification, provide examples of the present specification and describe technical features of the present specification together with the detailed description.

Hereinafter, the embodiments disclosed in this specification will be described in detail with reference to the accompanying drawings, but the same or similar elements are given the same reference numerals regardless of reference numerals, and redundant description thereof will be omitted.

In addition, in describing the embodiments disclosed in this specification, if it is determined that a detailed description of a related known technology may obscure the gist of the embodiment disclosed in this specification, the detailed description thereof will be omitted.

In addition, the accompanying drawings are only for easy understanding of the embodiments disclosed in this specification, the technical idea disclosed in this specification is not limited by the accompanying drawings, and all changes included in the spirit and technical scope of the present invention , it should be understood to include equivalents or substitutes.

Terms including ordinal numbers, such as first and second, may be used to describe various components, but the components are not limited by the terms. These terms are only used for the purpose of distinguishing one component from another.

Singular expressions include plural expressions unless the context clearly dictates otherwise.

In this application, terms such as "comprise" or "having" are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof external to the protection unit in the specification, but one or It should be understood that it does not preclude the possibility of the presence or addition of other features, numbers, steps, operations, components, parts, or combinations thereof beyond that.

Hereinafter, based on the above description, a detailed description of an anomaly detection system of a crane equipment using artificial intelligence according to a preferred embodiment of the present specification is as follows.

Hereinafter, the vibration signal and vibration data used in the present invention may be used as different terms for the same concept (electronic signal according to vibration) for convenience.

1 shows an anomaly detection system for crane equipment using artificial intelligence according to the present invention.

According to FIG. 1, the abnormal detection system of the crane equipment 1 using artificial intelligence according to the present invention includes a sensor 10 for detecting a vibration signal of the crane equipment 1 and a crane equipment 1 based on the vibration signal. It may include the computing device 100 that detects an abnormal state of .

The sensor 10 according to the present invention may be installed in the crane equipment 1. The sensor 10 according to the present invention may be a sensor 10 for detecting vibration occurring in the crane equipment 1. The sensor 10 according to the present invention may be for detecting a vibration signal according to an operation in the crane equipment 1. The vibration signal may be vibration generated by moving the joint of the crane device 1 or vibration generated by a motor included in the crane device 1. At this time, the joint part may include all parts accompanying movement that generates friction during the operation of the crane equipment 1. Therefore, the sensor 10 according to the present invention may be installed around a joint or a motor.

The computing device 100 according to the present invention may be configured to learn pre-input data and classify the data received from the sensor 10 based on the learning result. The computing device 100 according to the present invention may learn pre-input normal vibration data and abnormal vibration data. At this time, the normal vibration data may be stored as data of the vibration signal of the crane equipment 1 labeled as normal in advance. In addition, the abnormal vibration data may be stored as data of vibration signals of the crane equipment 1 previously labeled as abnormal. The computing device 100 according to the present invention may detect an abnormal state of the crane equipment 1 based on the learned result value and the vibration signal received from the sensor 10.

An abnormal state according to the present invention may include a case where a screw and / or a bolt is missing from the crane equipment 1, a case where a crack occurs, and a case where an abnormality occurs in the motor. The computing device 100 according to the present invention may learn the abnormal vibration data according to the above case by learning previously input data.

The computing device 100 according to the present invention may classify the abnormal state into a plurality of modes. In this case, the plurality of modes may be predetermined modes based on whether coping methods according to abnormal conditions are the same.

For example, a countermeasure for missing screws and/or bolts is to newly mount the missing screws and/or bolts, and one abnormal state may be defined based on this. In addition, a countermeasure in case a crack occurs in the crane equipment 1 or the part 2 included therein is to replace the cracked part 2, and another abnormal state may be defined based on this. In addition, a countermeasure for the case where an abnormality occurs in the motor is to check the fuel/engine oil or the motor itself, and another abnormal state may be defined based on this. As such, a plurality of modes according to the present invention may mean a mode predetermined based on whether a coping method according to an abnormal state is the same.

The computing device 100 according to the present invention may apply FFT (Fast Fourier Transform) to the steady vibration data and the abnormal vibration data during learning, and derive a first difference between the steady vibration data and the abnormal vibration data to which the FFT is applied. . The computing device 100 according to the present invention may learn the derived first difference value and detect an abnormal state of the crane equipment 1 based on this.

Specifically, the computing device 100 according to the present invention may extract a first vibration signal for a first period and a second vibration signal for a second period from vibration signals. At this time, the first vibration signal and the second vibration signal may be mechanically divided by a predetermined period. In addition, the first vibration signal and the second vibration signal may be divided according to the following process.

For example, the first vibration signal may be a vibration signal detected from the crane equipment 1 for a certain period of time. The second vibration signal may include a signal from when an amplitude greater than or equal to a predetermined magnitude is detected among the vibration signals to a predetermined period. In addition, the second vibration signal may include a signal from when a frequency equal to or greater than a predetermined magnitude is detected to a predetermined period.

The computing device 100 according to the present invention may apply FFT to the first vibration signal and the second vibration signal extracted as above. The computing device 100 according to the present invention may derive a second difference between the first vibration signal and the second vibration signal to which FFT is applied. The computing device 100 according to the present invention may detect an abnormal state of the crane equipment 1 based on the similarity between the second difference value and the first difference value. The computing device 100 according to the present invention may detect an abnormal state based on the similarity between the second difference value and the first difference value. The degree of similarity according to the present invention may be determined according to the K-center clustering process to be described later.

The computing device 100 according to the present invention may perform a K-center clustering process on the first difference value and the second difference value in order to derive a degree of similarity for the first difference value and the second difference value.

In addition, the computing device 100 according to the present invention finally generates two clusters based on the K-center clustering process, can detect an anomaly based on the Euclidean distance between the two clusters, and calculates the Euclidean distance Based on this, the types of abnormal conditions can be classified.

For example, the computing device 100 according to the present invention derives a degree of similarity between the second difference value and the first difference value, and when the degree of similarity is greater than a predetermined value, the crane equipment 1 may be classified as an abnormal state. . Elements extracted from the first and second difference value data (in the form of a wavelength) in order to derive a degree of similarity may be an amplitude and a time (s) at which the amplitude was measured.

The computing device 100 according to the present invention can learn and recognize this solid line shape. The computing device 100 according to the present invention derives the degree of similarity between the second difference value and the first difference value based on the amplitude extracted from the vibration signal and the time (s) at which the amplitude is measured, and may be classified according to the degree of similarity. . The degree of similarity may be set by the user in some cases.

For example, the computing device 100 according to the present invention may perform similarity classification of the first difference value and the second difference value based on the K-center clustering process. Specific details are described later.

2 illustrates a computing device in accordance with the present invention.

According to FIG. 2 , a computing device 100 according to the present invention may include a processor 110 , a memory 120 and a communication module 130 .

The processor 110 may execute commands stored in the memory 120 to control other components. The processor 110 may execute instructions stored in the memory 120 .

The processor 110 is a component capable of performing calculations and controlling other devices. Mainly, it may mean a central processing unit (CPU), an application processor (AP), a graphics processing unit (GPU), and the like. Also, the CPU, AP, or GPU may include one or more cores therein, and the CPU, AP, or GPU may operate using an operating voltage and a clock signal. However, while a CPU or AP consists of a few cores optimized for serial processing, a GPU may consist of thousands of smaller and more efficient cores designed for parallel processing.

The processor 110 may provide or process appropriate information or functions to a user by processing signals, data, information, etc. input or output through the components described above or by running an application program stored in the memory 120.

The memory 120 stores data supporting various functions of the control unit 100 . The memory 120 may store a plurality of application programs (application programs or applications) driven by the controller 100, data for operation of the controller 100, and commands. At least some of these application programs may be downloaded from the external controller 100 through wireless communication. In addition, the application program may be stored in the memory 120, installed in the control unit 100, and driven by the processor 110 to perform the operation (or function) of the control unit 100.

The memory 120 may be a flash memory type, a hard disk type, a solid state disk type, a silicon disk drive type, or a multimedia card micro type. ), card-type memory (eg SD or XD memory, etc.), RAM (random access memory; RAM), SRAM (static random access memory), ROM (read-only memory; ROM), EEPROM (electrically erasable programmable read -only memory), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, and an optical disk. Also, the memory 120 may include a web storage performing a storage function on the Internet.

The communication module 130 transmits and receives information with a base station or a camera having a communication function through an antenna. The communication module 130 may include a modulation unit, a demodulation unit, a signal processing unit, and the like.

Wireless communication may refer to communication using a wireless communication network using a communication facility previously installed by telecommunication companies and a frequency of the communication facility. At this time, the communication module 130 is CDMA (code division multiple access), FDMA (frequency division multiple access), TDMA (time division multiple access), OFDMA (orthogonal frequency division multiple access), SC-FDMA (single carrier frequency division multiple access), and the like, as well as the communication module 130 can also be used for 3rd generation partnership project (3GPP) long term evolution (LTE). In addition, not only 5G communication, which is currently being commercialized, but also 6G, which is scheduled to be commercialized later, can be used. However, the present specification may utilize a pre-installed communication network without being bound by such a wireless communication method.

In addition, as a short range communication technology, Bluetooth, BLE (Bluetooth Low Energy), Beacon, RFID (Radio Frequency Identification), NFC (Near Field Communication), infrared communication (Infrared Data Association; IrDA) ), UWB (Ultra Wideband), ZigBee, etc. may be used.

3 shows a functional configuration included in the memory according to the present invention, and FIG. 4 shows a graph to which the K-center clustering process performance module according to the present invention is applied.

According to FIG. 3, the functional components included in the memory (120 in FIG. 2, hereinafter the same) according to the present invention include an FFT application module 121, a difference value generation module 122, a CNN learning module 123, and a K-center A clustering process execution module 124 and a data storage module 125 may be included.

The data storage module 125 according to the present invention may be a module that stores all big data to be learning objects according to the present invention. Other functional components included in the memory 120 according to the present invention may receive big data from the data storage module 125 . The data storage module 125 according to the present invention may be a module that stores big data labeled by a user.

The FFT application module 121 according to the present invention may be a module that applies FFT to vibration signals or vibration data. FFT is a Fast Fourier Transform, which can be used to calculate an approximate value of a function. By applying the FFT, the learning speed can be improved.

The difference value generation module 122 according to the present invention may be a module that derives a difference value of vibration signals or vibration data to which FFT is applied. In order to derive the difference value, the difference value generating module 122 according to the present invention may set a reference time, list two vibration signals or vibration data based on the set reference time, and then derive a difference value. The reference time may be a preset time. Also, the reference time may be a time at which each vibration signal and vibration data start to be sensed.

The CNN learning module 123 according to the present invention may be a module for learning big data learned in the data storage module. Specifically, the CNN learning module 123 according to the present invention may learn a first difference value derived by FFT-applied steady vibration data and abnormal vibration data. Elements extracted from the first and second difference data (in the form of wavelengths) through the CNN learning module 123 according to the present invention coordinate the amplitude and the time (s) at which the amplitude was measured, and learn them . The CNN learning module 123 used at this time may be a module that implements a previously published CNN learning algorithm.

The K-center clustering process performing module 124 according to the present invention may be a module that measures and classifies similarities between the above-described first difference value and the above-described second difference value. Specifically, the K-center clustering process performing module 124 may generate K clusters and classify data by determining similarities based on distance values of the generated clusters.

The K-center clustering process refers to clustering given data into K clusters. In this K-center clustering process, after randomly selecting a sample value from a given data sample, the distance from the sample value to other sample data is measured. At this time, the nearest centroid is selected from each data sample, and the distance from the corresponding sample to another data sample value is calculated again. By repeating this process, clustering is achieved. That is, a plurality of clusters are created. This clustering technique can be mainly implemented in Python. Clustering is being used as one of the algorithms that can be used in signal analysis.

According to FIG. 4, as a result of performing the K-center clustering process, it is shown that measurement values are classified into clustering 1 and clustering 2. In this case, the data subject to clustering may be values obtained by coordinates of the amplitudes extracted from the first difference value and the second difference value and the time (s) at which the amplitude was measured.

That is, the K-center clustering process module according to the present invention coordinates the amplitude extracted from the first difference value and the second difference value and the time (s) at which the amplitude was measured, and performs clustering on the coordinated data. can

The K-center clustering process performance module according to the present invention calculates the first central value of clustering 1 (clustering for the first difference value) through the average calculation method, and calculates the second central value of clustering 2 (clustering for the second difference value). The center value can be calculated through the average calculation method. The K-center clustering process module according to the present invention calculates a final distance value between the first center value and the second center value, and when the size of the final distance value is greater than a predetermined value, the similarity of each cluster is determined in advance. It can be judged to be lower than the value.

That is, the K-center clustering process module according to the present invention calculates the final distance value of the first center value and the second center value, and when the size of the final distance value is greater than a predetermined value, the crane equipment is in an abnormal state. can be judged to have occurred. In addition, according to the type of the first difference value, which is a criterion for generating the abnormal state, the second difference value according to the present invention may be classified according to the mode according to the abnormal state.

According to FIG. 4, as an example, the K-center clustering process performing module according to the present invention sets the frequency (hz) according to the first difference value as feature 1, and sets the amplitude value as feature 2 to coordinate. . Since vibration is measured as an electrical signal by the sensor 10, feature 2 according to the present invention can be replaced with a voltage value (mV).

The K-center clustering process performing module according to the present invention may perform the K-center clustering process by setting feature 1 as the x-axis coordinate and feature 2 as the y-axis coordinate. The K-center clustering process performing module according to the present invention may calculate a first center value of cluster 1 and a second center value of cluster 2, and calculate a Euclidean distance between the first center value and the second center value. The Euclidean distance may be defined by Equation 1 below.

[Equation 1]

(However, d is a distance value, (x, y) is a coordinate value, x corresponds to the frequency (hz), and y corresponds to the measured voltage value (mV))

The system for detecting anomaly of crane equipment using artificial intelligence according to the present invention extracts the final distance value of cluster 1 and cluster 2, and can detect the abnormal state of the corresponding crane equipment based on the final distance value.

At this time, the final distance value according to the present invention can be extracted in the following order.

(1) Extracting the first average value (x1, y1) of cluster 1 and the second average value (x2, y2) of cluster 2

(2) The first average value and the second average value are coordinate values, respectively, and the final distance value (L) between the first average value and the second average value is extracted based on Equation 2 below.

[Equation 2]

(However, (x1, y1) is a coordinate value generated based on the average value of cluster 1 (= first central value), and (x2, y2) is generated based on the average value of cluster 2 (= second central value) coordinates)

In the present invention, when the extracted final distance value (L) is greater than a preset distance value, it may be defined that an abnormal state of the crane equipment is detected. In addition, the abnormality detection system of crane equipment using artificial intelligence according to the present invention may establish different criteria depending on the type of abnormal state or the location where the sensor 10 is installed. In general, when an abnormality occurs in a crane device, such as when a screw and/or bolt is missing, when a crack occurs, and when an abnormality occurs in a motor, the final distance value (L) is multiplied by a correction factor, and It is more efficient to compare with a preset distance value as a basis. This is referred to as the corrected final distance value (L') and can be extracted based on Equations 3 and 4 below.

[Equation 3]

(However, (x1, y1) is a coordinate value generated based on the average value of cluster 1, and (x2, y2) is a coordinate value generated based on the average value of cluster 2)

[Equation 4]

At this time, α is a correction constant for more accurate determination of the abnormal state, and the unit of α can be ignored. In this case, the temperature may be a value measured by a separately installed temperature sensor.

As such, when the fire occurrence information is generated based on the corrected final distance value (L') using Equations 3 and 4, there is an effect that the abnormal state of the crane equipment can be more accurately derived.

[Experiment example]

Accuracy of the system for detecting the abnormal state of the crane equipment based on the Euclidean distance generated using the correction constant according to the present invention and the system for detecting the abnormal state of the crane equipment based on the Euclidean distance without using the correction constant The comparison results are shown in Table 1 below.

In this case, the accuracy is calculated for 10 experts in the field, and may be a value calculated by performing experiments on a total of 100 cases with 10 experts in the same environment. Accuracy may be a comparison between an abnormal state detected by the system according to the present invention and an actual abnormal state.

Correction factor α not used Use correction factor α accuracy 78 91

(Unit: %) Table 1 shows the accuracy of cases according to the present invention. As a result of the experiment, the case where the correction coefficient α is used has significantly higher accuracy than the case where the correction coefficient α is not used.

5 is an example of comparing Euclidean distance values for each mode according to the present invention.

According to FIG. 5, when the computing device (100 in FIG. 1, hereinafter the same) according to the present invention detects an abnormal state of the crane equipment, a K-center clustering process is performed on the first difference value and the second difference value, , a plurality of clusters may be generated based on the performed K-center clustering process, and types of anomalies may be classified based on the generated plurality of clusters, an example of which may be the table in FIG. 5 . At this time, the plurality of clusters may finally be two clusters.

That is, the computing device 100 according to the present invention finally generates two clusters based on the K-center clustering process, can detect an anomaly based on the Euclidean distance between the two clusters, and calculates the Euclidean distance Based on this, the types of abnormal conditions can be classified.

As described above, the computing device 100 according to the present invention determines the ideal state based on the Euclidean distance between the average value of the first difference values (first central value) and the average value of the second difference values (second central value). detected. More specifically, the computing device 100 according to the present invention may classify the type of the second difference value according to the type of the first difference value.

For example, the first difference value according to the present invention may be data generated based on a first mode among abnormal states. The first mode may refer to a case where screws and/or bolts are missing. In addition, the first difference value according to the present invention may be data generated based on the second mode among abnormal conditions, and the second mode may mean a case where a crack occurs in the crane equipment. In addition, the first difference value according to the present invention may be data generated based on a third mode among abnormal conditions, and the third mode means a case in which an abnormality occurs in a motor that drives a part of crane equipment to operate. can do.

As such, the computing device 100 according to the present invention may derive the Euclidean distance between the first difference value generated for each mode and the second difference value measured and generated by the sensor 10 . The computing device 100 according to the present invention may classify the second difference value according to the Euclidean distance for each mode.

According to FIG. 5 , the first difference value may be divided into first to third modes for each mode. The second difference has a Euclidean distance of 25 from the first difference in the first mode, a Euclidean distance of 42 from the first difference in the second mode, and a Euclidean distance of 32 from the first difference in the third mode. has a Euclidean distance. That is, the second difference value has the closest Euclidean distance to the first difference value of the first mode, and the abnormal state of the crane equipment according to the present invention will be classified as corresponding to the case where the first mode screw and / or bolt is missing. can However, it is assumed that the first difference value and the second difference value are greater than the preset value in any case, and accordingly, the crane equipment according to the present invention is detected as an abnormal state.

In other words, if the Euclidean distance of the first difference value and the second difference value according to the present invention is greater than a predetermined value, the crane equipment is detected as an abnormal state. On this premise, when the crane equipment according to the present invention is detected in an abnormal state, the computing device 100 according to the present invention derives the Euclidean distance for each mode and classifies the abnormal state into the closest mode.

6 to 9 schematically show an abnormal state detection process according to the present invention.

According to FIG. 6, the abnormal state according to the present invention can be divided into a mode for screw detachment and a mode for cracks, and these classifications can be divided according to whether or not the countermeasures are the same. The computing device 100 according to the present invention may learn vibration data for each mode to create a learning module for each mode.

7 and 8, vibration data to which FFT is applied according to the present invention can generate a difference value (normalized FFT signal) obtained by subtracting abnormal vibration data from normal vibration data, and the computing device 100 according to the present invention difference can be learned.

According to FIG. 8 , the AI model included in the computing device 100 according to the present invention can be learned using multiclass decision forest, multiclass decision jungle, multiclass logistic regression, and multiclass neural network.

According to FIG. 9, the computing device 100 according to the present invention can classify whether a specific vibration signal corresponds to an abnormal state based on previously learned vibration data, and if so, which classification mode it corresponds to. there is.

10 schematically illustrates the configuration of a sensor according to the present invention.

According to FIG. 10 , the sensor 10 according to the present invention may include a body 11, a vibration receiving unit 12, a converter unit 13, a connection unit 14, and a coating layer 15.

The body according to the present invention includes the converter unit 13 and the connection unit 14 therein, and although not shown, may include a plurality of terminals electrically connected to the connection unit. The converter unit 13 according to the present invention may be connected to the vibration receiving unit 12 and the connection unit 14.

The vibration receiving unit 12 may be located closely to the vibration part and receive the vibration of the target. The converter unit 13 converts the vibration received from the vibration receiver 12 into an electrical signal and transmits it to the connection unit 14 . The connection unit 14 may transmit the converted electrical signal to an external device and receive necessary power.

The coating layer 15 is formed to surround the body and the vibration receiving unit, and may be configured to protect the body 11 and the vibration receiving unit 12 according to the present invention from external contamination and external impact.

The sensor 10 according to the present invention is installed in a crane equipment (1 in FIG. 1), and the crane equipment is for transporting a construction site or container cargo and has a high probability of being exposed to moisture with a lot of external dust or high salt concentration. big.

Accordingly, the present invention proposes a coating layer 15 for coating the sensor 10 to solve this problem.

Preferably, the sensor 10 may include an acrylic compound represented by Formula 1 below on its surface; It may be coated with a coating composition containing an organic solvent, inorganic particles and a dispersant.

[Formula 1]

here,

n and m are the same as or different from each other, each independently represent an integer of 1 to 100, and L ₁ is a biphenylene group.

When the sensor 10 is coated with the coating composition, it can exhibit excellent water repellency and stain resistance, so even if the sensor 10 installed in the external environment is exposed to a polluted environment and salt for a long time, it can sense clean and clear vibration data or vibration signals. can do.

The inorganic particles may be selected from the group consisting of silica, alumina, and mixtures thereof. The average diameter of the inorganic particles is 70 to 100 μm, but is not limited to the above examples. After the inorganic particles are formed as the coating layer 15 on the surface of the sensor 10, physical strength can be improved and moldability can be improved by maintaining the viscosity within a certain range.

The organic solvent is selected from the group consisting of methyl ethyl ketone (MEK), toluene, and mixtures thereof, and preferably methyl ethyl ketone may be used, but is not limited to the above examples.

As the dispersing agent, a polyester-based dispersing agent may be used, and specifically, TEGO-Disperse 670 (manufacturer : EVONIK) can be used, but it is not limited to the above examples, and all dispersants obvious to those skilled in the art can be used without limitation.

The coating composition may further include a stabilizer as other additives, and the stabilizer may include a UV absorber, an antioxidant, and the like, but is not limited to the above examples and may be used without limitation.

For forming the coating layer 15, the coating composition is more specifically an acrylic compound represented by Formula 1; organic solvents, inorganic particles and dispersants.

The coating composition may include 40 to 60 parts by weight of the acrylic compound represented by Formula 1, 20 to 40 parts by weight of inorganic particles, and 5 to 15 parts by weight of a dispersant, based on 100 parts by weight of the organic solvent. In the case of the above range, a synergistic effect is expressed to the extent that the water repellency effect due to the interaction of each component is of critical significance, and when it is out of the above range, the synergistic effect is rapidly reduced or almost nonexistent.

More preferably, the viscosity of the coating composition is 1500 to 1800 cP, and when the viscosity is less than 1500 cP, when applied to the surface of the sensor 10, there is a problem that it is not easy to form the coating layer 15 because it flows down, and the coating composition exceeds 1800 cP. In the case of doing so, there is a problem in that formation of a uniform coating layer 15 is not easy.

[Preparation Example 1: Preparation of coating layer]

1. Preparation of coating composition

A coating composition was prepared by mixing methyl ethyl ketone with an acrylic compound represented by Formula 1, inorganic particles, and a dispersant:

[Formula 1]

here,

n and m are the same as or different from each other, each independently represent an integer of 1 to 100, and L ₁ is a biphenylene group.

A more specific composition of the antistatic composition is shown in Table 2 below.

TX1 TX2 TX3 TX4 TX5 organic solvent 100 100 100 100 100 acrylic compound 30 40 50 60 70 inorganic particles 10 20 30 40 50 dispersant One 5 10 15 20

(Unit weight part) 2. Preparation of coating layer

After coating the coating composition of DX1 to DX5 on one surface of the sensor 10, the coating layer 15 was formed by curing.

[Experimental Example]

1. Evaluation of surface appearance

Due to the difference in viscosity of the coating composition, after the coating layer 15 was prepared, a sensory evaluation was conducted on whether or not a uniform surface was formed. Evaluation was conducted on whether or not a uniform coating layer 15 was formed, and the evaluation was conducted according to the following criteria.

○: uniform coating layer formation

×: Formation of non-uniform coating layer

TX1 TX2 TX3 TX4 TX5 sensory evaluation Х ○ ○ ○ Х

When forming the coating layer 15, when the viscosity is less than a certain amount, flow occurs on the surface of the sensor 10, and it is difficult to form a uniform coating layer 15 after the curing process in many cases. Accordingly, a problem of lowering the production yield may occur. In addition, even when the viscosity is too high, it is difficult to uniformly apply the composition and it is impossible to form a uniform coating layer 15 .

2. Measurement of water repellency angle

After forming the coating layer 15 on the surface of the sensor 10, the results of measuring the water repellency angle are shown in Table 4 below.

Advancing contact angle (“) Static contact angle (“) receding contact angle (“) TX1 117.1±2.9 112.1±4.1 < 10 TX2 132.4±1.5 131.5±2.7 141.7±3.4 TX3 138.9±3.0 138.9±2.7 139.8±3.7 TX4 136.9±2.0 135.6±2.6 140.4±3.4 TX5 116.9±0.7 115.4±3.0 < 10

As shown in Table 4, after forming the coating layer 15 using the coating compositions of TX1 to TX5, the result of measuring the contact angle was confirmed. TX1 and TX5 measured receding contact angles less than 10 degrees. That is, it was confirmed that a phenomenon in which water droplets are pinned occurs when the coating composition is out of the optimal range for preparing the coating composition. On the other hand, it was confirmed that the pinning phenomenon did not occur in TX2 to 4, indicating that excellent waterproofing effect could be exhibited.

3. Fouling resistance evaluation

The sensor 10 having the coating layer 15 according to the above embodiment was attached to crane equipment and exposed to the seaside environment for 40 days. As a comparative example (Con), the same sensor 10 without the coating layer 15 was used, and each sensor 10 was attached to the same position among crane equipment.

After that, the degree of contamination and corrosion of the sensor 10 before and after the experiment was evaluated as related, and for objective comparison, the result was compared with a comparative example in which the coating layer 15 was not formed, and the result was evaluated by an index of 1 to 10, and the following Table 5 shows. In the index below, the lower the number, the better the stain resistance.

Con TX1 TX2 TX3 TX4 TX5 stain resistance 10 7 3 3 3 8

(Unit: Index) Referring to Table 5 above, when the coating layer 15 is formed on the sensor 10, while installing the sensor 10 in the external environment, even if the sensor 10 is exposed to the outside, high contamination resistance is maintained for a long time. It can be seen that vibration data can be collected in a form that is easy to analyze for a period. In particular, in the case of TX2 to TX4, it can be confirmed that the contamination resistance by the coating layer 15 is very excellent.

Hereinafter, based on the above description, a method for detecting an abnormality of crane equipment using artificial intelligence according to another preferred embodiment of the present specification will be described in detail as follows.

A subject performing the abnormal detection method of crane equipment using artificial intelligence according to the present invention may be the above-described computing device 100 according to the present invention or a processor included in the computing device 100.

In addition, in the description of the method for detecting abnormality of crane equipment using artificial intelligence according to the present invention, the same or overlapping contents as the above may be omitted, which is clearly understood by those skilled in the art of the present invention. It can be.

11 and 12 show a method for detecting abnormalities in crane equipment using artificial intelligence according to the present invention.

According to FIG. 11, the abnormal detection method of crane equipment using artificial intelligence according to the present invention includes the steps of learning normal vibration data generated from crane equipment in a normal state and abnormal vibration data generated from crane equipment in an abnormal state (S1100 ), receiving a vibration signal generated from the target crane equipment (S1200) and detecting an abnormal state of the target crane equipment based on the learned result value and vibration signal (S1300).

At this time, the normal vibration data may be vibration data generated by the crane equipment determined to be in a normal state, and the abnormal vibration data may be vibration data generated by the crane equipment determined to be in an abnormal state, and the target crane equipment may be in an abnormal state. As the equipment to be sensed of, it may mean a crane equipment to which the sensor 10 is attached. The sensor 10 according to the present invention measures normal vibration data and abnormal vibration data at the same location, and may be installed at a location corresponding to the corresponding location among target crane equipment. This can be equally understood in the system according to an embodiment of the present invention described above.

According to FIG. 12, the learning step (S1100) according to the present invention is a step of applying FFT (Fast Fourier Transform) to steady vibration data and abnormal vibration data (S1110), and to the steady vibration data and abnormal vibration data to which the FFT is applied. It may include deriving a first difference value (S1120) and learning the first difference value (S1130). At this time, as described in an embodiment of the present invention, the first difference value according to the present invention may vary according to the type of abnormal state.

The above-described present invention can be modeled as a computer readable code on a medium on which a program is recorded. The computer-readable medium includes all types of recording devices in which data readable by a computer system is stored. Examples of computer-readable media include Hard Disk Drive (HDD), Solid State Disk (SSD), Silicon Disk Drive (SDD), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. , and also includes those modeled in the form of carrier waves (eg, transmission over the Internet). Accordingly, the above detailed description should not be construed as limiting in all respects and should be considered illustrative. The scope of this specification should be determined by reasonable interpretation of the appended claims, and all changes within the equivalent scope of this specification are included in the scope of this specification.

Any or other embodiments of the present invention described above are not mutually exclusive or distinct. Certain or other embodiments of the present invention described above may be used in combination or combination of respective configurations or functions.

The above detailed description should not be construed as limiting in all respects and should be considered illustrative. The scope of the present invention should be determined by reasonable interpretation of the appended claims, and all changes within the equivalent scope of the present invention are included in the scope of the present invention.

1: Crane equipment
2: Part
10: sensor
100: computing device

Claims (7)

In a system for detecting anomalies of crane equipment that generates vibration signals according to operation,
a sensor that detects the vibration signal; and
A computing device that learns pre-input normal vibration data and abnormal vibration data, and detects an abnormal state of the crane equipment based on the learned result value and the vibration signal;
Crane equipment abnormality detection system using artificial intelligence.
According to claim 1,
The computing device classifies the abnormal state into a plurality of modes,
The plurality of modes are predetermined based on whether the coping methods according to the abnormal state are the same,
Crane equipment abnormality detection system using artificial intelligence.
According to claim 1,
The computing device,
Applying FFT (Fast Fourier Transform) to the steady vibration data and abnormal vibration data, deriving a first difference value to the steady vibration data and abnormal vibration data to which the FFT is applied, and learning the first difference value,
Crane equipment abnormality detection system using artificial intelligence.
According to claim 3,
The computing device,
Extracting a first vibration signal for a first period and a second vibration signal for a second period with respect to the vibration signal;
The computing device,
FFT is applied to the first vibration signal and the second vibration signal, a second difference value between the first vibration signal and the second vibration signal to which the FFT is applied is derived, and the second difference value and the first difference value To detect the abnormal state of the crane equipment based on the similarity of,
Crane equipment abnormality detection system using artificial intelligence.
According to claim 4,
The first difference value is data according to the type of the abnormal state,
The computing device,
When an abnormal state of the crane equipment is detected, a K-center clustering process is performed on the first difference value and the second difference value, and finally two clusters are generated based on the K-center clustering process, Deriving the similarity based on the Euclidean distance between the two clusters,
Crane equipment abnormality detection system using artificial intelligence.
Learning normal vibration data generated from crane equipment in a normal state and abnormal vibration data generated from crane equipment in an abnormal state;
Receiving a vibration signal generated from the target crane equipment; and
Detecting an abnormal state of the target crane equipment based on the learned result value and the vibration signal; Including,
The learning step is
applying a Fast Fourier Transform (FFT) to the steady vibration data and the abnormal vibration data;
deriving a first difference between the normal vibration data and the abnormal vibration data to which the FFT is applied; and
Comprising the step of learning the first difference value,
A method for detecting anomalies in crane equipment using artificial intelligence.
A non-transitory computer readable medium storing instructions,
The instructions, when executed by a processor, cause the processor to perform the method of claim 6.

Images