A Defect Detection Method and System for Circuit Breaker Contacts

Abstract

The present application provides a method and system for defect detection of a circuit breaker contact, relating to the technical field of circuit breaker detection, including: in response to detecting that the circuit breaker contact to be detected starts to be powered on, obtaining a thermal distribution map of the contact surface of the circuit breaker contact to be detected once every preset time length to obtain a list P of thermal distribution maps; obtaining a list K of defect point coordinates of the circuit breaker contact to be detected according to the preset ultrasonic detection method; and inputting P into a spatio-temporal self-attention mechanism image detection model to obtain a detection result corresponding to the circuit breaker contact to be detected. By combining the method of detecting the thermal map through power-on and the ultrasonic detection method, the finally output image recognition result, that is, the defect detection result, is more accurate than using only the ultrasonic detection method or the method of detecting the thermal map through power-on alone.

Description

Background Art

[0002] During the operation of the power system, as a key control and protection device, the reliability of the performance of the circuit breaker is directly related to the stability and safety of the entire power system. For a relatively large circuit breaker, the circuit breaker contact, as a core component, the existence of internal defects will lead to an increase in contact resistance, increased heating, and even cause equipment failures and power accidents. Therefore, it is of great significance to accurately and efficiently detect the internal defects of the circuit breaker contacts of relatively large circuit breakers.

[0003] Currently, for the detection of internal defects of circuit breaker contacts of relatively large circuit breakers, mainly two technical means are adopted: ultrasonic detection and obtaining a thermal map by energization. Ultrasonic detection uses the principle that ultrasonic waves will reflect, refract, and scatter when encountering an interface during propagation in different media to judge internal defects. However, this technology has obvious defects. For example, for the circuit breaker contacts of relatively large circuit breakers with complex structures, ultrasonic waves will encounter multiple reflection interfaces during propagation, which is likely to generate interference signals, making it difficult for the detector to distinguish the real defect signals from the interference signals, reducing the accuracy and reliability of defect detection. And obtaining a thermal map by energization is to energize the circuit breaker contact to generate heat, and use an infrared thermal imager to obtain the thermal map of the contact surface, and judge internal defects according to the temperature distribution in the thermal map. But this technology also has many deficiencies. On the one hand, this method mainly reflects the temperature distribution on the contact surface. For deep internal defects of the contact, due to heat diffusion and heat loss during the heat transfer process, the heat generated by the internal defects may not be fully reflected on the surface, resulting in difficulty in accurately detecting internal defects. On the other hand, the detection results are significantly affected by environmental factors. For example, environmental temperature, humidity, wind speed, etc. will all interfere with the temperature measurement on the contact surface. In high-temperature, high-humidity or strong-wind environments, the measurement error increases, and misjudgment or missed judgment may occur. In addition, for the circuit breaker contacts of relatively large circuit breakers, due to their large heat dissipation area and relatively dispersed thermal distribution, the temperature anomaly area is not obvious enough, further increasing the difficulty of defect detection.

[0004] In summary, the existing ultrasonic detection and obtaining a thermal map by energization detection technologies have problems such as low detection accuracy and being greatly affected by external factors when detecting the internal defects of circuit breaker contacts of relatively large circuit breakers, and cannot meet the actual needs of the power system for the reliability detection of circuit breaker contacts. There is an urgent need to develop a more efficient and accurate detection technology. Summary of the Invention

[0005] In view of the above technical problems, the present application provides a method and system for defect detection of circuit breaker contacts, which at least partially solves the problems existing in the prior art.

[0006] In the first aspect of the present application, a method for defect detection of a circuit breaker contact is provided, and the method includes:

[0007] S100, in response to detecting that the circuit breaker contact to be detected starts to be energized, obtain a thermal distribution map of the contact surface of the circuit breaker contact to be detected at each preset time interval to obtain a list of thermal distribution maps P = (P1, P2,..., P i ,..., P n ); i = 1, 2,..., n; where n is the number of thermal distribution maps obtained; P i is the map identifier of the i-th thermal distribution map obtained;

[0008] S200, according to the preset ultrasonic detection method, obtain a list of defect point coordinates K = (K1, K2,..., K j ,..., K m ); j = 1, 2,..., m; where m is the number of defect points determined according to the preset ultrasonic detection method; K j is the coordinate of the j-th defect point determined according to the preset ultrasonic detection method in the preset three-dimensional rectangular coordinate system; K j = (x j , y j , z j ); where x j , y j , z j are the coordinate values of the j-th defect point on the x-axis, y-axis and z-axis respectively;

[0009] S300, input P into the spatio-temporal self-attention mechanism image detection model to obtain the detection result corresponding to the circuit breaker contact to be detected; where the time self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate value of each defect point in the z-axis in K; the spatial self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in the x-axis and y-axis in K; the detection result is used to characterize the position of the defect of the circuit breaker contact to be detected.

[0010] In the second aspect of the present application, a defect detection system for a circuit breaker contact is provided, and the system includes:

[0011] A heat map acquisition unit, configured to, in response to detecting that the circuit breaker contact to be detected starts to be energized, obtain a thermal distribution map of the contact surface of the circuit breaker contact to be detected at each preset time interval to obtain a list of thermal distribution maps P = (P1, P2,..., P i ,..., P n ); i = 1, 2,..., n; where n is the number of thermal distribution maps obtained; P iThe graph identifier of the i-th thermal distribution graph obtained;

[0012] A coordinate list acquisition unit, configured to obtain a list K=(K1, K2, …, K j , …, K m ) of the coordinates of the defect points of the breaker contact to be detected according to a preset ultrasonic detection method; j = 1, 2, …, m; where m is the number of defect points determined according to the preset ultrasonic detection method; K j is the coordinate of the j-th defect point determined according to the preset ultrasonic detection method in a preset three-dimensional rectangular coordinate system; K j =(x j , y j , z j ); where x j , y j , z j are respectively the coordinate values of the j-th defect point on the x-axis, y-axis and z-axis;

[0013] A detection unit, configured to input P into a spatio-temporal self-attention mechanism image detection model to obtain a detection result corresponding to the breaker contact to be detected; wherein, the temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in K on the z-axis; the spatial self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in K on the x-axis and y-axis; the detection result is used to characterize the position of the defect of the breaker contact to be detected.

[0014] The present application has at least the following beneficial effects:

[0015] This application first obtains the thermal distribution map of the contact surface of the contact of the circuit breaker to be detected based on the energization acquisition thermal map detection method, and starts energizing from the contact of the circuit breaker to be detected. Then, once every preset time length, a thermal distribution map list is obtained. Here, since defects may be located at different depth positions of the contact of the circuit breaker to be detected, when it is heated by energization, for the contacts of the circuit breaker to be detected at different depth positions, their appearance degrees in the thermal map of the contact surface of the contact of the circuit breaker to be detected may be different after different time lengths of starting energization. Here, the depth refers to the height of the defect from the contact surface. Due to the existence of the above differences, this application sequentially obtains an image sequence at preset intervals starting from the contact of the circuit breaker to be detected. After that, according to the preset ultrasonic detection method, a list of defect point coordinates of the contact of the circuit breaker to be detected is obtained. Here, according to the preset ultrasonic detection method, a list of coordinates of all suspected defect positions inside the contact of the circuit breaker to be detected is initially obtained. To further improve the accuracy of defect detection, this application inputs P into the spatio-temporal self-attention mechanism image detection model. The temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate value of each defect point on the z-axis in the defect point coordinate list; the spatial self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point on the x-axis and y-axis in the defect point coordinate list. Because the coordinate of each defect point on the z-axis in the defect point coordinate list represents the depth information of the initially determined defect point, and since the temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model includes several temporal attention weights, the temporal self-attention mechanism weights the image features at different time steps (different image acquisition times) through these weights for weighted fusion. The temporal self-attention mechanism can adaptively allocate the importance of image information at different time steps according to the content and task requirements of the image sequence. And here, the temporal feature of the image (for the contacts of the circuit breaker to be detected at different depth positions, their appearance degrees in the thermal map of the contact surface of the contact of the circuit breaker to be detected may be different after different time lengths of starting energization) initially reflects the thermal map display situation of defects in different depth positions. The temporal self-attention matrix is adjusted according to the coordinate of each defect point on the z-axis in the defect point coordinate list, that is, the temporal attention weights are adjusted, and at the same time, the depth information of the defect points obtained from the thermal map and the depth information of the defect points obtained from the ultrasonic detection method are combined, so that the model can determine the key attention images in the images corresponding to the time sequence. Similarly, based on the coordinate values of each defect point on the x-axis and y-axis in the defect point coordinate list, the spatial attention weights in the spatial self-attention matrix are adjusted, so that the model can assign different importance levels to different regions of each picture, so as to facilitate the model to focus on the important regions in different regions of each picture.In summary, by combining the method of obtaining a thermal image through energization detection and the ultrasonic wave detection method, the final output image recognition result, that is, the defect detection result, is more accurate than using only the ultrasonic wave detection method or the method of obtaining a thermal image through energization detection alone. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0017] Figure 1 It is a flowchart of a defect detection method for a circuit breaker contact provided by an embodiment of the present application;

[0018] Figure 2 It is a structural block diagram of a defect detection system for a circuit breaker contact provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0020] It should be noted that the terms "first", "second", etc. in the description and claims of the present application and the above drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units does not necessarily need to be limited to those clearly listed steps or units, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0021] Note that the following description relates to various aspects of embodiments within the scope of the appended claims. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this application, those skilled in the art should understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of the aspects described herein can be used to implement an apparatus and / or practice a method. Additionally, this apparatus can be implemented and this method can be practiced using other structures and / or functionality in addition to one or more of the aspects described herein.

[0022] Please refer to Figure 1 As shown, an embodiment of the present application provides a method for defect detection of a circuit breaker contact, and the method includes;

[0023] S100, in response to detecting that the circuit breaker contact to be detected starts to be energized, obtain a thermal distribution map of the contact surface of the circuit breaker contact to be detected at each preset time interval to obtain a list of thermal distribution maps P = (P1, P2,..., P i ,..., P n ); i = 1, 2,..., n; where n is the number of thermal distribution maps obtained; P i is the map identifier of the i-th thermal distribution map obtained.

[0024] Specifically, since defects may be located at different depth positions of the circuit breaker contact to be detected, and when it is heated by energization, for the circuit breaker contact to be detected at different depth positions, the degree of appearance in the thermal map of the contact surface of the circuit breaker contact to be detected may be different after different time lengths of starting to be energized. Here, the depth refers to the height of the defect from the contact surface. Due to the existence of the above differences, in this embodiment, an image sequence is obtained at preset intervals starting from when the circuit breaker contact to be detected starts to be energized. As an example: A defect located at a relatively deep position may become more obvious in the thermal map corresponding to the contact surface of the circuit breaker contact to be detected 10 minutes after starting to be energized, while a defect located at a relatively shallow position may become more obvious in the thermal map corresponding to the contact surface of the circuit breaker contact to be detected 20 minutes after starting to be energized.

[0025] S200, according to a preset ultrasonic detection method, obtain a list of defect point coordinates K = (K1, K2,..., K j ,..., K m ); j = 1, 2,..., m; where m is the number of defect points determined according to the preset ultrasonic detection method; K jThe coordinates of the j-th defect point determined according to the preset ultrasonic detection method in the preset three-dimensional rectangular coordinate system; K j = (x j , y j , z j ); where x j , y j , z j are the coordinate values of the j-th defect point on the x-axis, y-axis, and z-axis, respectively.

[0026] Specifically, according to the preset ultrasonic detection method, a list of defect point coordinates of the circuit breaker contact to be detected is obtained. Here, according to the preset ultrasonic detection method, a list of coordinates of all suspected defect positions inside the circuit breaker contact to be detected is initially obtained.

[0027] Among them, step S200 includes:

[0028] S210, determine a number of detection points and a detection path according to the structure of the circuit breaker contact to be detected.

[0029] S220, sequentially detect each detection point according to the detection path to obtain the ultrasonic signal corresponding to each detection point.

[0030] S230, analyze the ultrasonic signal corresponding to each detection point to obtain the list of defect point coordinates K of the circuit breaker contact to be detected.

[0031] In one embodiment, ultrasonic waves are used to detect internal defects in relatively large circuit breaker contacts. This is mainly based on the principle that when ultrasonic waves propagate in different media and encounter an interface, reflection, refraction, and scattering will occur. By analyzing the signals of the reflected waves, it is determined whether there are internal defects. First, according to the material, size, and detection requirements of the circuit breaker contacts, an ultrasonic detector with an appropriate frequency and power is selected. Usually, the frequency is between 1 - 10 MHz to ensure that the ultrasonic waves can effectively penetrate the contacts and obtain clear reflected signals. At the same time, corresponding ultrasonic probes, such as straight probes, inclined probes, etc., are prepared to adapt to different detection positions and angles. The surface of the contacts is cleaned to remove impurities such as oil, dust, and rust to ensure good coupling between the ultrasonic probe and the contact surface and reduce signal attenuation. A suitable coupling agent, such as vaseline, paste, water, etc., is selected and applied to the contact surface and the contact part of the probe so that the ultrasonic waves can smoothly enter the interior of the contacts. According to the structure of the contacts and the possible defective parts, the detection points and detection paths are planned. Generally, multiple detection points are set at different positions and in different directions of the contacts to form a three-dimensional detection space grid to comprehensively cover the internal area of the contacts. For example, for cylindrical contacts, detection points can be evenly arranged in the axial and circumferential directions; for flat contacts, detection points can be set at a certain interval on the plane and layered detection can be carried out in the thickness direction. The ultrasonic probe is closely attached to the detection points on the contact surface through the coupling agent, and the probe is moved along the predetermined detection path. During the movement, the ultrasonic detector emits ultrasonic pulses and receives the signals reflected from the interior of the contacts. These signals are displayed on the screen of the detector in the form of waveforms, and the detection personnel observe the changes in the waveforms in real time and record the characteristics of the reflected signals at different positions. Normally, when ultrasonic waves propagate in a uniform medium, the waveform of the reflected wave is relatively stable and regular. When there are defects inside the contacts, such as pores, cracks, inclusions, etc., it will cause the propagation path of the ultrasonic waves to change, and parameters such as the amplitude, phase, and time delay of the reflected wave will also change. For example, when encountering a crack, the amplitude of the reflected wave will increase significantly and multiple reflection peaks may appear; when there are pores, the amplitude of the reflected wave will decrease and the waveform will become blurred. According to the time delay of the reflected wave and the propagation speed of the ultrasonic waves in the contact material, the distance between the defect and the detection surface can be calculated, thereby determining the position of the defect inside the contacts. Combining the results of different detection points and detection paths, three-dimensional spatial positioning of the defects can be carried out to depict the approximate shape and distribution range of the defects inside the contacts. The waveform data, defect position information, and relevant detection parameters obtained during the detection process are recorded in detail. The coordinates of each detection point, the characteristic parameters of the reflected wave, the defect judgment results, etc. can be recorded in tabular form. At the same time, image recording software can be used to capture screenshots of the waveforms for subsequent analysis and comparison. Finally, the three-dimensional coordinates of each initially determined suspected defect point can be obtained.

[0032] In S300, input P into the spatio-temporal self-attention mechanism image detection model to obtain the detection result corresponding to the circuit breaker contact to be detected. Among them, the temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate value of each defect point in the z-axis in K; the spatial self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in the x-axis and y-axis in K; the detection result is used to characterize the position of the defect of the circuit breaker contact to be detected.

[0033] Specifically, input P into the spatio-temporal self-attention mechanism image detection model. Here, the temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate value of each defect point in the z-axis in the defect point coordinate list; the spatial self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in the x-axis and y-axis in the defect point coordinate list. This is because the coordinate of each defect point in the z-axis in the defect point coordinate list represents the depth information of the preliminarily determined defect point. And since the temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model includes several temporal attention weights, the temporal self-attention mechanism weights the image features at different time steps (different image acquisition times) through these weights for weighted fusion. The temporal self-attention mechanism can adaptively allocate the importance of image information at different time steps according to the content of the image sequence and the task requirements. And here, the temporal feature of the image (the degree to which the circuit breaker contacts at different depth positions appear in the heat map of the contact surface of the circuit breaker contact to be detected after different lengths of time of starting to energize may be different) preliminarily reflects the heat map display situation of the defect at different depth positions. Adjust the temporal self-attention matrix according to the coordinate of each defect point in the z-axis in the defect point coordinate list, that is, adjust the temporal self-attention weights, and at the same time combine the depth information of the defect point obtained from the heat map and the depth information of the defect point obtained from the ultrasonic detection method, so that the model can determine the key attention image in the images corresponding to the time series. Similarly, based on the coordinate values of each defect point in the x-axis and y-axis in the defect point coordinate list, adjust the spatial self-attention weights in the spatial self-attention matrix, so that the model can assign different importance levels to different regions of each picture, so as to facilitate the model to focus on the important regions in different regions of each picture. In summary, through the combination of the heat map detection method obtained by energization and the ultrasonic wave detection method in this application, the final output picture recognition result, that is, the defect detection result is more accurate than using only the ultrasonic detection method or the heat map detection method obtained by energization alone.

[0034] In an exemplary embodiment of the present application, the temporal self-attention matrix is determined according to the following steps:

[0035] In S310, obtain the initial temporal self-attention matrix A, where A meets the following conditions:

[0036] ;

[0037] where a i，i is the initial temporal attention weight of the i-th heat map with respect to the i-th heat map; a i，n is the initial temporal attention weight of the i-th heat map with respect to the n-th heat map.

[0038] Specifically, the initial temporal attention weight represents the degree of association or importance between the current image and the images at other time steps at a specific time step. The larger a i，n is, the more important the model considers image i for understanding and processing the information of image n at the current time step. The temporal self-attention mechanism weights and fuses the image features at different time steps through these weights, thereby obtaining a new feature representation that integrates temporal sequence information. This enables the model to fully utilize the information at other relevant time steps when processing the current image, capture the long-term dependencies in the image sequence, and improve the accuracy and comprehensiveness of feature extraction. Larger attention weights will make the model pay more attention to the corresponding image information, thereby highlighting the key information and important features in the image sequence, which helps the model perform better in recognition tasks. Relatively smaller attention weights mean that the model considers the corresponding image information less important for the processing at the current time step, thereby suppressing the influence of these possible noises or irrelevant information on the model output to a certain extent. In this way, the model can focus more on the information related to the task and improve the robustness and accuracy of the model.

[0039] S320. Adjust the mapping table according to K and the preset time to obtain the temporal adjustment coefficient matrix AT, where AT meets the following conditions:

[0040] ;

[0041] where at i，i is the total adjustment coefficient corresponding to the initial temporal attention weight of the i-th heat map with respect to the i-th heat map; at i，i = at i，i，1 × at i，i，2 ×... × at i，i，j ×... at i，i，m ; where at i，i，j is the adjustment coefficient of the j-th defect point with respect to a i，i ; if the coordinate value of the j-th defect point on the z-axis is included in the temporal adjustment target list corresponding to the preset time adjustment mapping table and contains a i，i , then at i，i，j = α; α > 1; otherwise, at i，i，j= 1; The preset time adjustment mapping table contains several coordinate values of the z-axis and a list of time adjustment targets corresponding to each coordinate value of the z-axis; The list of time adjustment targets includes several time adjustment targets; The time adjustment target is any initial time attention weight in A.

[0042] Specifically, the coordinate value of each defect point on the z-axis determined by the ultrasonic detection method is used to adjust the initial time attention weight. Since there may be more than one defect point determined by the ultrasonic detection method, then, for the same initial time attention weight, it may be adjusted multiple times. That is, the initial time attention weight that is adjusted multiple times may be more important.

[0043] S330, according to A and AT, obtain the time self-attention matrix AM, where AM meets the following conditions:

[0044] ;

[0045] where, am i，i is the time attention weight of the i-th heat distribution map with respect to the i-th heat distribution map; am i，i meets the following conditions: am i，i = a i，i × at i，i .

[0046] It can be understood that the preset time adjustment mapping table contains several coordinate values of the z-axis and a list of time adjustment targets corresponding to each coordinate value of the z-axis, that is, for the current circuit breaker contact to be detected, the adjustment targets are different for defect points at different depths (several coordinate values of the z-axis). Here, the adjustment target is the important heat distribution map in the heat distribution maps obtained at different times mapped by the defect points. Therefore, the list of time adjustment targets corresponding to a certain coordinate value of the z-axis is the initial time attention weight (images of certain time series) that the defect points at this depth need to focus on. After determining the time adjustment target, the adjustment coefficient of each adjustment target is determined as α greater than 1, and for the time attention weight that is not the time adjustment target, it is not adjusted. This embodiment increases the value of the initial time attention weight that needs to be focused on, that is, enables the model to focus on the time series images corresponding to each time adjustment target.

[0047] In an exemplary embodiment of the present application, the spatial self-attention matrix is determined according to the following steps:

[0048] S340, obtain the initial spatial self-attention matrix list B = (B1, B2,..., B i ,..., B n ); B i is the initial spatial self-attention matrix corresponding to the i-th heat distribution map;

[0049] where, Bi Meet the following conditions:

[0050] ;

[0051] p = 1, 2, …, q; where q is the number of pixel points included in each heat distribution map; b i，p，p is the spatial attention weight of the p-th pixel point to the p-th pixel point in the i-th heat distribution map.

[0052] Specifically, the spatial self-attention matrix focuses on the relationship between spatial positions within a single image. Since each image has its unique spatial structure and features, it is necessary to calculate the spatial self-attention matrix of each image separately to capture the importance distribution of different positions within the image. p = 1, 2, …, q; where q = q H ×q S ; q H represents the number of pixel points included in the image in the vertical direction; q S represents the number of pixel points included in the image in the horizontal direction; for the spatial self-attention matrix, calculate the attention weights between spatial positions within each image.

[0053] S350, according to K and the preset spatial adjustment mapping table, obtain the spatial adjustment coefficient matrix list BT = (BT1, BT2, …, BT i , …, BT n ); BT i is the adjustment coefficient matrix corresponding to B i ; where BT i meets the following conditions: ;

[0054] where bt i，p，p is the total adjustment coefficient corresponding to the initial spatial attention weight of the p-th pixel point to the p-th pixel point in the i-th heat distribution map; bt i，p，p = bt i，p，p1 × bt i，p，p，2 × … × bt i，p，p，j × … bt i，p，p，m ; bt i，p，p，j is the adjustment coefficient of the j-th defect point to bt i，p，p ; if the coordinate values of the j-th defect point on the x-axis and y-axis are included in the spatial adjustment target list corresponding to the preset spatial adjustment mapping table and contain b i，p，p , then bt i，p，p = α; α > 1; otherwise, bt i，p，p= 1; The preset spatial adjustment mapping table contains several pairs of x-axis and y-axis coordinate values, and a list of spatial adjustment targets corresponding to each pair of x-axis and y-axis coordinate values; The list of spatial adjustment targets includes several spatial adjustment targets; The spatial adjustment target is any initial spatial attention weight in B.

[0055] S360, according to B and BT, obtain the list of spatial self-attention matrices BM=(BM1, BM2,..., BM i ,..., BM n ); BM i is the spatial self-attention matrix corresponding to B i ; Among them, BM i meets the following conditions: ;

[0056] Among them, bm i，p，p is the spatial attention weight of the p-th pixel point to the p-th pixel point in the i-th heat distribution map; bm i，p，p meets the following conditions: bm i，p，p = b i，p，p × bt i，p，p .

[0057] Specifically, the coordinate values of each defect point determined by the ultrasonic detection method on the x-axis and y-axis are used to adjust the initial spatial attention weight. Since there may be more than one defect point determined by the ultrasonic detection method, then, the same initial spatial attention weight may be adjusted multiple times. That is, the initial time attention weight that is adjusted multiple times may be relatively important.

[0058] It can be understood that the preset spatial adjustment mapping table contains several coordinate values of the x-axis and y-axis, and a list of spatial adjustment targets corresponding to each coordinate value of the x-axis and y-axis. That is, for the current circuit breaker contact to be detected, the adjustment targets corresponding to defect points at different positions (several coordinate values of the x-axis and y-axis) are different. Here, the list of adjustment targets corresponding to a certain coordinate value of the x-axis and y-axis is the initial spatial attention weight (the area that needs to be focused on in a certain image) that the defect points at this position need to focus on. After determining the spatial adjustment target, the adjustment coefficient of each spatial adjustment target is determined to be α greater than 1, and for the time attention weight that is not a spatial adjustment target, it is not adjusted. This embodiment increases the value of the initial spatial attention weight that needs to be focused on, that is, enables the model to focus on the key areas in the image corresponding to each spatial adjustment target.

[0059] It should be noted that α is a parameter, not a hyperparameter, and can be obtained by training the spatio-temporal self-attention mechanism image detection model.

[0060] In an exemplary embodiment of the present application, when obtaining the thermal distribution map of the contact surface of the breaker contact to be detected, a constant current or a non-constant current can be used. The non-constant current can be any one of a fluctuating current, a continuously increasing current, a continuously decreasing current, etc. In one embodiment, the defect points at different depths are more obvious in the thermal map obtained using a non-constant current.

[0061] Please refer to Figure 2 As shown, an embodiment of the present application provides a defect detection system 100 for a breaker contact, and the system includes:

[0062] A heat map acquisition unit 110, configured to obtain a thermal distribution map of the contact surface of the breaker contact to be detected at each preset time interval in response to detecting that the breaker contact to be detected starts to be energized, so as to obtain a thermal distribution map list P = (P1, P2,..., P i ,..., P n ); i = 1, 2,..., n; where n is the number of acquired thermal distribution maps; P i is the map identifier of the i-th acquired thermal distribution map.

[0063] A coordinate list acquisition unit 120, configured to obtain a defect point coordinate list K = (K1, K2,..., K j ,..., K m ) according to a preset ultrasonic detection method; j = 1, 2,..., m; where m is the number of defect points determined according to the preset ultrasonic detection method; K j is the coordinate of the j-th defect point determined according to the preset ultrasonic detection method in a preset three-dimensional rectangular coordinate system; K j = (x j , y j , z j ); where x j , y j , z j are the coordinate values of the j-th defect point on the x-axis, y-axis, and z-axis, respectively.

[0064] A detection unit 130, configured to input P into a spatio-temporal self-attention mechanism image detection model to obtain a detection result corresponding to the breaker contact to be detected; where the temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in the z-axis in K; the spatial self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in the x-axis and y-axis in K; the detection result is used to characterize the position of the defect of the breaker contact to be detected.

[0065] Embodiments of the present application also provide a computer program product, which includes program code. When the program product runs on an electronic device, the program code is used to cause the electronic device to execute the steps in the methods according to various exemplary embodiments of the present application described above in this specification.

[0066] In addition, although the steps of the methods in the present application are described in a specific order in the drawings, this does not require or imply that these steps must be executed in that specific order, or that all the steps shown must be executed to achieve the desired result. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step for execution, and / or one step may be decomposed into multiple steps for execution, etc.

[0067] Through the description of the above embodiments, those skilled in the art can easily understand that the exemplary embodiments described herein can be implemented by software, or by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, server, mobile terminal, or network device, etc.) to execute the methods according to the embodiments of the present application.

[0068] In an exemplary embodiment of the present application, an electronic device capable of implementing the above method is also provided.

[0069] Those skilled in the art can understand that various aspects of the present application can be implemented as a system, method, or program product. Therefore, various aspects of the present application can be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation combining hardware and software aspects, which can be collectively referred to herein as "circuit", "module", or "system".

[0070] The electronic device according to this embodiment of the present application. The electronic device is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0071] The electronic device is presented in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: the at least one processor described above, the at least one storage described above, and a bus connecting different system components (including the storage and the processor).

[0072] Among them, the storage stores program code, and the program code can be executed by the processor, so that the processor executes the steps according to various exemplary embodiments of the present application described in the above "Exemplary Methods" section of this specification.

[0073] The memory may include a readable medium in the form of volatile memory, such as random access memory (RAM) and / or cache memory, and may further include read-only memory (ROM).

[0074] The memory may also include a program / utility having a set (at least one) of program modules, such program modules including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may include an implementation of a network environment.

[0075] The bus may represent one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus structures.

[0076] The electronic device may also communicate with one or more external devices (such as a keyboard, a pointing device, a Bluetooth device, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device, and / or may communicate with any device that enables the electronic device to communicate with one or more other computing devices (such as a router, a modem, etc.). Such communication may be through an input / output (I / O) interface. Also, the electronic device may communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter. As shown in the figure, the network adapter communicates with other modules of the electronic device through the bus. It should be understood that although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0077] Through the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or can be implemented by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which may be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which may be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the embodiments of the present application.

[0078] In an exemplary embodiment of the present application, a computer-readable storage medium is further provided, on which a program product capable of implementing the above methods of this specification is stored. In some possible implementation manners, various aspects of the present application can also be implemented in the form of a program product, which includes program code. When the program product runs on a terminal device, the program code is used to cause the terminal device to execute the steps according to various exemplary embodiments of the present application described in the above "Exemplary Method" section of this specification.

[0079] The program product can adopt any combination of one or more readable media. The readable media can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0080] The computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries the readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable signal medium can also be any readable medium other than the readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device.

[0081] The program code contained on the readable medium can be transmitted by any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0082] The program code for performing the operations of this application can be written in any combination of one or more programming languages, including object-oriented programming languages such as Java, C++, etc., and also including conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, executed as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user's computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., by using an Internet service provider to connect through the Internet).

[0083] In addition, the above-mentioned drawings are only schematic illustrations of the processes included in the method according to the exemplary embodiments of this application, rather than for restrictive purposes. It is easy to understand that the processes shown in the above-mentioned drawings do not indicate or limit the chronological order of these processes. Additionally, it is also easy to understand that these processes can be executed synchronously or asynchronously, for example, in multiple modules.

[0084] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, such a division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more of the above-mentioned modules or units can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0085] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in this application should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A method for defect detection of a circuit breaker contact, characterized in that Including: S100, in response to detecting that the contact of the circuit breaker to be detected starts to be energized, obtain a thermal distribution map of the contact surface of the contact of the circuit breaker to be detected once every preset time length, so as to obtain a list of thermal distribution maps P = (P1, P2, …, P i , …, P n ); i = 1, 2, …, n; where n is the number of obtained thermal distribution maps; P i is the map identifier of the i-th obtained thermal distribution map; S200, according to the preset ultrasonic detection method, obtain the list of coordinates of defect points of the circuit breaker contact to be detected, K = (K1, K2, …, K j , …, K m ); j = 1, 2, …, m; where m is the number of defect points determined according to the preset ultrasonic detection method; K j is the coordinate of the j-th defect point determined according to the preset ultrasonic detection method in the preset three-dimensional rectangular coordinate system; K j = (x j , y j , z j ); where x j , y j , z j are the coordinate values of the j-th defect point on the x-axis, y-axis, and z-axis respectively; S300, input P into the spatio-temporal self-attention mechanism image detection model to obtain the detection result corresponding to the circuit breaker contact to be detected; wherein, the temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate value of each defect point in the z-axis in K; the spatial self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in the x-axis and y-axis in K; the detection result is used to characterize the position of the defect of the circuit breaker contact to be detected.

2. The defect detection method for the breaker contact according to claim 1, characterized in that, The temporal self-attention matrix is determined according to the following steps: S310, obtain the initial temporal self-attention matrix A, where A meets the following conditions: ； Among them, a i，i is the initial time attention weight of the i-th heat distribution map for the i-th heat distribution map; S320, according to K and the preset time adjustment mapping table, obtain the time adjustment coefficient matrix AT, where AT meets the following conditions: ； Among them, at i，i is the total adjustment coefficient corresponding to the initial time attention weight of the i-th thermal distribution map for the i-th thermal distribution map; at i，i = at i，i，1 × at i，i，2 × … × at i，i，j × … at i，i，m ; Among them, at i，i，j is the adjustment coefficient of the j-th defect point for a i，i ; If the coordinate value of the j-th defect point on the z-axis is included in the time adjustment target list corresponding to the preset time adjustment mapping table for a i，i , then at i，i，j = α; α > 1; Otherwise, at i，i，j = 1; The preset time adjustment mapping table includes several coordinate values of the z-axis and the time adjustment target list corresponding to each coordinate value of the z-axis; The time adjustment target list includes several time adjustment targets; The time adjustment target is any initial time attention weight in A; S330, according to A and AT, obtain the temporal self-attention matrix AM, where AM meets the following conditions: ； Among them, am i，i is the time attention weight of the i-th thermal distribution map with respect to the i-th thermal distribution map; am i，i meets the following conditions: am i，i =a i，i ×at i，i .

3. The defect detection method for the breaker contact according to claim 1, characterized in that, The spatial self-attention matrix is determined according to the following steps: S340, obtain the initial spatial self-attention matrix list B = (B1, B2, …, B i , …, B n ); B i is the initial spatial self-attention matrix corresponding to the i-th heat distribution map; Among them, B i meets the following conditions: ； p = 1, 2, …, q; q is the number of pixel points included in each heat distribution map; b i，p，p is the spatial attention weight of the p-th pixel point to the p-th pixel point in the i-th heat distribution map; S350, adjust the mapping table according to K and the preset space to obtain a list of space adjustment coefficient matrices BT=(BT1, BT2, …, BT i , …, BT n ); among them, BT i is the adjustment coefficient matrix corresponding to B i ; BT i meets the following conditions: ； Among them, bt i，p，p is the total adjustment coefficient corresponding to the initial spatial attention weight of the p-th pixel point to the p-th pixel point in the i-th thermal distribution map; bt i，p，p =bt i，p，p1 ×bt i，p，p，2 ×…×bt i，p，p，j ×…bt i，p，p，m ; bt i，p，p，j is the adjustment coefficient of the j-th defect point to bt i，p，p ; if the coordinate values of the j-th defect point on the x-axis and y-axis are included in the spatial adjustment target list corresponding to the preset spatial adjustment mapping table as b i，p，p , then bt i，p，p =α; α>1; otherwise, bt i，p，p =1; the preset spatial adjustment mapping table includes several pairs of x-axis and y-axis coordinate values, and the spatial adjustment target list corresponding to each pair of x-axis and y-axis coordinate values; the spatial adjustment target list includes several spatial adjustment targets; the spatial adjustment target is any initial spatial attention weight in B; S360, obtain a list of spatial self-attention matrices BM=(BM1, BM2, …, BM i , …, BM n ); BM i is the corresponding spatial self-attention matrix of B i ; where BM i meets the following conditions: ； Among them, bm i，p，p is the spatial attention weight of the p-th pixel point to the p-th pixel point in the i-th heat distribution map; bm i，p，p meets the following conditions: bm i，p，p = b i，p，p × bt i，p，p .

4. The defect detection method for the breaker contact according to any one of claims 2-3, characterized in that, α is obtained by training the spatio-temporal self-attention mechanism image detection model.

5. The defect detection method for the breaker contact according to claim 1, characterized in that, When obtaining the thermal distribution map of the contact surface of the circuit breaker contact to be detected, a constant current is used.

6. The defect detection method for the breaker contact according to claim 1, wherein When obtaining the thermal distribution map of the contact surface of the circuit breaker contact to be detected, a non-constant current is used.

7. The defect detection method for the breaker contact according to claim 1, characterized in that, Step S200 includes: S210, according to the structure of the circuit breaker contact to be detected, determine a number of detection points and a detection path; S220, sequentially detect each detection point according to the detection path to obtain the ultrasonic signal corresponding to each detection point; S230, analyze the ultrasonic signal corresponding to each detection point to obtain the defect point coordinate list K of the circuit breaker contact to be detected.

8. A defect detection system for a circuit breaker contact, characterized in that, The system includes: A heat map acquisition unit, which is configured to, in response to detecting that the contact of the circuit breaker to be detected starts to be energized, acquire a heat distribution map of the contact surface of the contact of the circuit breaker to be detected once every preset time length, so as to obtain a heat distribution map list P = (P1, P2,..., P i ,..., P n ); i = 1, 2,..., n; where n is the number of acquired heat distribution maps; P i is the map identifier of the i-th acquired heat distribution map; A coordinate list acquisition unit, configured to obtain a coordinate list K=(K1, K2, …, K j , …, K m ) of the defect points of the breaker contact to be detected according to a preset ultrasonic detection method; j = 1, 2, …, m; where m is the number of defect points determined according to the preset ultrasonic detection method; K j is the coordinate of the j-th defect point determined according to the preset ultrasonic detection method in a preset three-dimensional rectangular coordinate system; K j =(x j , y j , z j ); where x j , y j , z j are the coordinate values of the j-th defect point on the x-axis, y-axis, and z-axis, respectively; A detection unit, configured to input P into the spatio-temporal self-attention mechanism image detection model to obtain the detection result corresponding to the circuit breaker contact to be detected; wherein, the temporal self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate value of each defect point in the z-axis in K; the spatial self-attention matrix in the spatio-temporal self-attention mechanism image detection model is determined according to the coordinate values of each defect point in the x-axis and y-axis in K; the detection result is used to characterize the position of the defect of the circuit breaker contact to be detected.

Images