An intelligent safety monitoring system and method for railway pantograph network

Abstract

The present application provides an intelligent safety monitoring system and method for a railway pantograph network, comprising a monitoring data module, a monitoring scheme module, an abnormality assessment module, and an abnormality handling module; the monitoring data module comprises an acquisition module, a processing module, and an analysis module; the abnormality judgment module is capable of judging whether an image abnormality occurs through an image evaluation value; the monitoring scheme module comprises a normal state monitoring scheme employed when no image abnormality occurs and an abnormality assessment monitoring scheme employed when an image abnormality occurs; the abnormality assessment module is provided with an abnormality assessment period; when the abnormality judgment module judges that an image abnormality occurs, the abnormality assessment module monitors the image during the abnormality assessment period using the abnormality assessment monitoring scheme and judges the abnormality degree; the abnormality handling module is configured to allocate processing steps corresponding to the abnormality degree according to the obtained abnormality degree. The present application improves monitoring safety and accuracy.

Description

Technical Field

[0001] The present application relates to the field of safety monitoring of railway pantograph networks, and in particular to an intelligent safety monitoring system and method for railway pantograph networks. Background Art

[0002] The train pantograph is a device that the train uses to obtain current from the contact network. Its carbon sliding strips are in direct contact with the contact network wires to obtain current for the train to use. During the high-speed operation of the train, due to factors such as vibration, changes in contact network height, contact pressure, and the quality of the pantograph material, the pantograph may have some safety problems, such as hooking foreign objects, deformation, missing bow angles, abnormal landing or electric sparks, etc., which will pose a threat to the safety of the train operation. Therefore, in order to promptly discover and deal with possible safety hazards and ensure the normal operation of the train, it is necessary to monitor the operation of the pantograph in real time. The current method of monitoring the pantograph is to collect pantograph images in real time, process and analyze the collected images, determine whether any pantograph abnormalities are found in the images, and evaluate the safety status of the pantograph through the results of data monitoring. When processing and analyzing images, the collected images are affected by environmental reasons, lighting reasons, shooting reasons and other comprehensive factors that affect the subsequent analysis and processing. Image processing methods generally include image cropping, preprocessing, edge detection, edge extraction, etc. Specifically, image cropping is to remove unnecessary parts such as surrounding scenery in the image, and only retain key areas for judgment. Cropping can be based on known shooting position and angle information; then the cropped image is preprocessed. Preprocessing includes grayscale (converting color images into grayscale images), filtering denoising (using Gaussian filtering, median filtering and other methods to remove noise in the image), contrast enhancement (increasing the contrast of the image through methods such as histogram equalization) to improve the accuracy of edge detection. Edge detection can identify areas of significant brightness variation in pantograph images, such as the edges of the contact wire and pantograph runner. Common edge detection algorithms include the Sobel operator and the Canny operator. These algorithms detect edges by calculating the gradient magnitude and direction of each pixel in the image. Furthermore, based on edge detection, detected edges can be further extracted and contour fitted to connect scattered edge points into complete contours, accurately locating the upper and lower edges of the contact wire and pantograph runner. Contour fitting can use the Hough transform (which maps edge points in image space to parameter space and finds corresponding lines or curves by detecting peaks in the parameter space) to fit the straight edges of the contact wire and pantograph runner. After image processing and analysis using these methods, the results are output as data for subsequent analysis and monitoring. However, since image capture is easily affected by the train operating environment, such as lighting fluctuations, background noise, and train speed, problems such as unclear images, inaccurate feature acquisition, and false alarms can occur.In addition, since real-time monitoring is required during the entire train operation, computing resources such as GPU, memory, and data transmission speed also limit the number of image frames that can be processed per second. Specifically, the higher the number of captured frames, the greater the consumption of computing resources. At the same time, the speed of railway vehicles will also affect the number of frames and recognition processing. When analyzing images, if the monitoring sampling is too frequent, the analysis and processing frame rate is too high, or even every frame is processed and analyzed, it will lead to excessive consumption of computing resources. Especially when the train is running at high speed, it is difficult to achieve completely real-time feedback and it is easy to trigger false alarms. If the sampling is insufficient, it is impossible to provide the large amount of labeled data required for model training, making the basic data values inaccurate and resulting in insufficient robustness of the electrical fault detection model. Summary of the Invention

[0003] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are for illustrative purposes only, and those skilled in the art will readily appreciate other obvious variations. The basic principles of the present invention defined in the following description may be applied to other embodiments, variations, improvements, equivalents, and other technical solutions that do not depart from the spirit and scope of the present invention.

[0004] In order to solve the technical problem, the present application provides an intelligent safety monitoring system for a railway pantograph network, which is characterized by comprising a monitoring and judgment module and an abnormality handling module;

[0005] Wherein, the monitoring and judging module includes a first judging module, a second judging module and a third judging module;

[0006] The first judgment module monitors the pantograph in a normal state monitoring scheme to obtain a first image, and judges whether the first image is abnormal according to an evaluation value of the first image;

[0007] When the first image is judged to be abnormal, the second judgment module monitors the pantograph according to the abnormality degree assessment monitoring scheme during the abnormality degree assessment period, obtains a second image, determines the abnormality classification of the second image, and determines the abnormality degree of the pantograph during the abnormality degree assessment period based on the evaluation value of the second image and the abnormality classification;

[0008] When the abnormality level of the pantograph during the abnormality level assessment period reaches the system warning standard, the abnormality processing module triggers the system abnormality warning and the monitoring ends;

[0009] When the abnormality level of the pantograph during the abnormality level assessment period does not reach the system warning standard, the third judgment module obtains a third image using a secondary assessment scheme during the secondary assessment period, determines the abnormality classification of the third image, and determines the abnormality level of the pantograph during the secondary assessment period based on the assessment value and the abnormality classification of the third image;

[0010] When the abnormality level of the secondary evaluation period reaches the system warning standard, the abnormality handling module triggers the system abnormality warning and the monitoring ends; when the abnormality level of the secondary evaluation period reaches the system normal standard, the first judgment module monitors the pantograph with the normal state monitoring scheme; when the abnormality level of the secondary evaluation period is between normal and warning, the second judgment module continues to monitor the pantograph with the abnormality level evaluation monitoring scheme.

[0011] The abnormality classifications of the second image and the third image include: normal, mild abnormality, and severe abnormality.

[0012] Among them, the second judgment module judges the abnormality classification according to the evaluation value of the second image, and then obtains the corresponding abnormality classification coefficient and abnormality distribution according to the abnormality classification of the second image, calculates the evaluation score of the abnormality degree evaluation period, and judges the abnormality degree of the pantograph in the abnormality degree evaluation period according to the evaluation score.

[0013] The system further includes an acquisition module, a processing module and an analysis module; wherein the acquisition module is used to acquire images of the pantograph during train operation; the processing module can process the acquired images to obtain key areas for judging the pantograph status; the analysis module can analyze the processed images to obtain an evaluation value representing the pantograph status.

[0014] The processing module processes the collected image in the following ways: cropping, preprocessing, edge detection and edge extraction.

[0015] The analysis module uses a deep learning model or a traditional machine vision algorithm to perform data analysis on the image processed by the processing module, thereby obtaining an evaluation value of the image.

[0016] The present application also provides a monitoring method using the above-mentioned intelligent safety monitoring system for the railway pantograph network, the steps comprising:

[0017] S1, set the monitoring time T i Get the i-th image U i , get U i The evaluation value Q i ;

[0018] According to the basic evaluation value Q0 of the image;

[0019] When Q i ≥Q0, the image U i If it is determined to be an abnormal image, the process goes to step S2;

[0020] S2, monitor the pantograph with the abnormality evaluation monitoring scheme E1 during the abnormality evaluation period △T1, and set a total of m1 images U, U=[U1, U2, U3, ..., U m1 ], where the jth image is U j , get image U j The evaluation value Q j ;

[0021] According to the basic evaluation value Q0, the evaluation interval threshold Q of the abnormality evaluation period is obtained max and Q min , where Q min ≤ln2 Q0;Q max ≥2Q0-Q min ;

[0022] When Q j ≥Q max When image U j It is judged to be a severe abnormality;

[0023] When Q j ≤Q min When the image U j Judged as normal;

[0024] When Q min <Q j <Q max When image U j Judgement of mild abnormality;

[0025] S3, among the m1 images acquired during the abnormality assessment period △T1, including c1 severely abnormal images and c2 slightly abnormal images, set the severe abnormality coefficient ω1 and the slightly abnormality coefficient ω2, where ω1>ω2;

[0026] Obtain the abnormal distribution p1 = (c1 + c2) / m1 during the abnormality evaluation period;

[0027] Among them, set the j1th severely abnormal image U j1 The evaluation value is Q j1 , the j2nd mildly abnormal image U j2 The evaluation value is Q j2 ;

[0028] Set the evaluation threshold score W0;

[0029] Obtain the evaluation score of the abnormality evaluation period △T1 ;

[0030] When W1≥W0, the pantograph is judged to be severely abnormal, the system alarms and intervenes;

[0031] When W1<W0, go to step S4;

[0032] S4, monitor the pantograph with the secondary evaluation scheme during the secondary evaluation period △T2, and set the zth image among the m2 images obtained as U z , image U z The evaluation value is Q z ; Assume that m2 images include d1 severely abnormal images and d2 slightly abnormal images, and obtain the abnormal distribution p2 = (d1 + d2) / m2 during the second evaluation period;

[0033] Among them, set the z1th severely abnormal image U z1 The evaluation value is Q z1 , the z2nd mildly abnormal image U z2 The evaluation value is Q z2 ;

[0034] Obtain assessment scores for the second assessment period ;

[0035] When W2≥W0, the pantograph is judged to be severely abnormal, the system alarms and intervenes;

[0036] When W2≤W1, the normal state monitoring scheme is restored;

[0037] When W1<W2<W0, the secondary evaluation plan is released and the monitoring plan is continued based on the abnormality level evaluation.

[0038] The beneficial effects achieved by monitoring this application are as follows:

[0039] By adopting appropriate monitoring methods and technical means, this application can timely discover and deal with potential safety hazards, ensure the normal operation of trains, reduce false alarms, rationally allocate resources, strengthen the collection of targeted endpoint data, improve the accuracy of deep learning models, and improve the intelligence, precision and efficiency of the monitoring system.

[0040] This application overcomes the problem that most mechanism models are only simplified linear systems. It can judge complex situations with nonlinearity, high degrees of freedom and multi-variable coupling. It has fast triggering speed, low cost, and very rapid start-up and recovery of early warning, and has a wide range of applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can also be obtained based on these drawings.

[0042] Figure 1 This is a flow chart of the steps of the monitoring method for the railway pantograph network of this application. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present application are clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present application. It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0044] In order to reduce false alarms, allocate resources rationally, strengthen the collection of targeted endpoint data, and improve the accuracy of deep learning models, the present application provides an intelligent safety monitoring system for a railway pantograph network, wherein the intelligent safety monitoring system includes a monitoring and judgment module and an abnormality handling module; wherein the monitoring and judgment module includes a first judgment module, a second judgment module and a third judgment module; wherein the first judgment module monitors the pantograph with a normal state monitoring scheme to obtain a first image, and judges whether the first image is abnormal according to the evaluation value of the first image; when the first image is judged to be abnormal, the second judgment module monitors the pantograph with an abnormality degree assessment monitoring scheme during an abnormality degree assessment period to obtain a second image, judges the abnormality classification of the second image, and judges the abnormality degree of the pantograph during the abnormality degree assessment period according to the evaluation value and abnormality classification of the second image; When the degree of abnormality of the pantograph during the abnormality assessment period reaches the system warning standard, the abnormality handling module triggers the system abnormality warning and monitoring ends. When the degree of abnormality of the pantograph during the abnormality assessment period does not reach the system warning standard, the third judgment module obtains a third image using the secondary assessment scheme during the secondary assessment period, determines the abnormality classification of the third image, and determines the degree of abnormality of the pantograph during the secondary assessment period based on the assessment value and abnormality classification of the third image. When the degree of abnormality during the secondary assessment period reaches the system warning standard, the abnormality handling module triggers the system abnormality warning and monitoring ends. When the degree of abnormality during the secondary assessment period reaches the system normal standard, the first judgment module monitors the pantograph using the normal state monitoring scheme. When the degree of abnormality during the secondary assessment period is between normal and warning, the second judgment module continues to monitor the pantograph using the abnormality assessment monitoring scheme. The abnormality classifications of the second and third images include normal, mild abnormality, and severe abnormality. Wherein, the second judgment module judges the abnormality classification based on the evaluation value of the second image, and then obtains the corresponding abnormality classification coefficient and abnormality distribution based on the abnormality classification of the second image, calculates the evaluation score of the abnormality degree evaluation period, and judges the abnormality degree of the pantograph in the abnormality degree evaluation period based on the evaluation score. Wherein, the system also includes an acquisition module, a processing module and an analysis module; wherein, the acquisition module is used to acquire images of the pantograph during train operation; the processing module is capable of processing the acquired images to obtain key areas that can be used to judge the state of the pantograph; the analysis module is capable of analyzing the processed images to obtain an evaluation value representing the state of the pantograph. wherein, the processing module processes the acquired images in the following steps: cropping, preprocessing, edge detection and edge extraction. wherein, the analysis module uses a deep learning model or a traditional machine vision algorithm to perform data analysis on the images processed by the processing module, thereby obtaining an evaluation value of the image.

[0045] Specifically, in the intelligent safety monitoring system, the acquisition module acquires monitoring images through a camera or other shooting device, and the processing module processes the acquired monitoring images to obtain the brightness conditions of key areas (edge areas of the pantograph and contact line); the processed images are uploaded to the cloud processor, and the processed images are analyzed by the analysis module using a deep learning model or other machine vision algorithm to obtain an image evaluation value representing the pantograph status.

[0046] In one embodiment, a basic evaluation value Q0 for the pantograph is set based on historical data of normal images (the brightness of the edge area of the pantograph and the contact wire when no electric shock occurs). When the evaluation value of the monitored image exceeds the basic evaluation value Q0, the image is judged to be abnormal (i.e., a possible electric shock may have occurred). Since brightness varies under different lighting conditions and speeds, the basic evaluation value Q0 can be adjusted based on different shooting times, lighting conditions, and train speeds.

[0047] When monitoring is performed under normal conditions, the monitoring time T i The i-th image U obtained i The evaluation value is Q i The characteristics μ0 used to evaluate the basic assessment value Q0 in the normal state monitoring scheme include image sampling frequency, number of image processing items, image analysis algorithm complexity, train speed, computing resources, and data transmission speed. When setting these parameters, it's important to consider that higher sampling frequencies and more processing items result in clearer image processing and more accurate results, but this increases processing time and computing resources, and also increases the data transmission speed requirements.

[0048] To ensure data accuracy, a network neural model was established to determine whether the basic evaluation value Q0 is appropriate for this monitoring and evaluation. Features such as image sampling frequency, image processing items, visual algorithm complexity, train speed, computing resources, and data transmission speed were put into the trained SVM model for analysis to determine whether the basic evaluation value is applicable under the current train operation conditions. The SVM algorithm is as follows:

[0049] Objective function:

[0050]

[0051] Where W and b are the plane coefficients, Represents the classification label of the sample, , is the training sample; W is the plane coefficient, when expanded to n-dimensional space, it is an n-dimensional vector such as: W=[W1,W2,...,Wn], is the transpose of W, ||W|| is the norm of the hyperplane;

[0052] Since the SVM objective function assumes that the data is linearly separable, but in reality there will be noisy data, slack variables and penalty parameters are added to increase the model tolerance through slack variables:

[0053]

[0054] in, is the penalty coefficient, is a slack variable, and the penalty term is determined by all outliers. The optimization problem is transformed into a dual problem using the Lagrange multiplier method and KKT conditions, and is solved using the SMO method; is the Lagrange multiplier; the dual problem obtained by mapping the model to a high dimension is:

[0055]

[0056] The kernel function selected in this paper is the Gaussian function:

[0057]

[0058] in, is the width function of the function;

[0059] Use SVM classifier to analyze whether it is applicable, -1 represents yes and 1 represents no.

[0060] When Q i ≥Q0, add the i-th image as an abnormal image to the abnormal image list, and i Set to the corresponding abnormal time point, Q i Set to the corresponding abnormal evaluation value;

[0061] Since an abnormal image is obtained, the abnormality evaluation period △T1 is entered at this time. During the abnormality evaluation period △T1, monitoring is performed using the abnormality evaluation monitoring scheme. Among them, the image processing coefficient of the abnormality evaluation monitoring scheme is ;

[0062] The abnormality evaluation monitoring scheme is used to obtain m1 images U within the abnormality evaluation period △T1, where U=[U1, U2, U3, …, U m ], where the jth image is U j , get image U j The evaluation value Q j ;

[0063] The evaluation threshold Q of the abnormality evaluation period is obtained according to the basic evaluation value Q0 max and Q min , where Q min≤ln2Q0;Q max ≥2Q0-Q min ;

[0064] When Q j ≥Q max When image U j It is judged to be a severe abnormality;

[0065] When Q j ≤Q min When image U j Judged as normal;

[0066] When Q min <Q j <Q max When image U j Judgement of mild abnormality;

[0067] According to the above judgment results, the m1 images in the abnormality assessment period △T1 include c1 severely abnormal images and c2 slightly abnormal images, and the abnormality distribution of the abnormality assessment period is obtained as p1 = (c1 + c2) / m1;

[0068] Among them, set the j1th severely abnormal image U j1 The evaluation value is Q j1 , the j2nd mildly abnormal image U j2 The evaluation value is Q j2 ;

[0069] Set the severe anomaly coefficient ω1 and the mild anomaly coefficient ω2, where ω1>ω2;

[0070] Set the evaluation base score W0;

[0071] Obtain the evaluation score of the abnormality evaluation period △T1 ;

[0072] When W1 ≥ W0, the pantograph network is judged to be abnormally large and an alarm is issued to relevant personnel;

[0073] When W1<W0, it enters the secondary evaluation period;

[0074] During the secondary evaluation period △T2, the image processing coefficient of the secondary evaluation scheme is monitored. ;

[0075] The second evaluation scheme is used to obtain m2 images in the second evaluation period △T2 (△T2<△T1), among which the zth image is U z , get image U z The evaluation value Q zThe m2 images acquired during the secondary evaluation period △T2 include d1 severely abnormal images and d2 slightly abnormal images, and the abnormal distribution of the secondary evaluation period is obtained as p2 = (d1 + d2) / m2;

[0076] Among them, set the z1th severely abnormal image U z1 The evaluation value is Q z1 , the z2nd mildly abnormal image U z2 The evaluation value is Q z2 ;

[0077] Obtain assessment scores for the second assessment period ;

[0078] When W2 ≥ W0, the pantograph network is judged to be abnormally large and an alarm is issued to relevant personnel;

[0079] When W2≤W1, the system returns to normal state for monitoring;

[0080] When W1<W2<W0, the secondary evaluation plan is released and monitoring is continued with the abnormality degree evaluation monitoring plan.

[0081] Alternatively, ultraviolet imaging can be used to acquire images. Since pantograph-catenary arcing emits a significant amount of ultraviolet radiation (200-400nm), which has a wavelength lower than that of visible light, sudden changes in the UV data can be used to identify the occurrence of pantograph-catenary arcing. However, UV sensors are expensive to install and can easily miss or falsely detect minor electrical faults. Therefore, a combination of UV imaging and photogrammetry can be used.

[0082] Furthermore, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein the method described in the above method embodiment is executed when the program is run.

[0083] Furthermore, the present invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the method described in the above method embodiment through the computer program.

[0084] Furthermore, it should be understood that since the configuration of each module is merely for the purpose of illustrating the functional units of the apparatus of the present invention, the physical devices corresponding to these modules may be the processor itself, or a portion of the software in the processor, a portion of the hardware, or a combination of software and hardware. Therefore, the number of modules in the figure is merely illustrative.

[0085] It will be understood by those skilled in the art that all or part of the processes in the methods for implementing the above embodiments of the present invention may also be accomplished by instructing related hardware through a computer program, and the computer program may be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of each of the above method embodiments may be implemented. The computer program includes computer program code, which may be in source code form, object code form, executable file, or some intermediate form. The computer-readable storage medium may include any entity or device, medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory, random access memory, electric carrier signal, telecommunication signal, and software distribution medium, etc., which may carry the computer program code.

[0086] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is implemented by a network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0087] Those skilled in the art will understand that the various modules in the device can be adaptively split or merged, and such splitting or merging of specific modules will not cause the technical solution to deviate from the principles of the present invention. Therefore, the technical solutions after splitting or merging will fall within the scope of protection of the present invention.

Claims

1. A monitoring method for an intelligent safety monitoring system of a railway pantograph network, characterized in that: The intelligent safety monitoring system of the railway pantograph network includes a monitoring and judgment module and an abnormality handling module; wherein the monitoring and judgment module includes a first judgment module, a second judgment module and a third judgment module; wherein the first judgment module monitors the pantograph with a normal state monitoring scheme to obtain a first image, and judges whether the first image is abnormal according to the evaluation value of the first image; when the first image is judged to be abnormal, the second judgment module monitors the pantograph with an abnormality degree assessment monitoring scheme during an abnormality degree assessment period to obtain a second image, judges the abnormality classification of the second image, and judges the abnormality degree of the pantograph during the abnormality degree assessment period according to the evaluation value and abnormality classification of the second image; When the abnormality level of the pantograph in the abnormality level assessment period reaches the system early warning standard, the abnormality handling module triggers the system abnormality early warning and the monitoring ends; when the abnormality level of the pantograph in the abnormality level assessment period does not reach the system early warning standard, the third judgment module obtains the third image in the secondary assessment period with the secondary assessment scheme, judges the abnormality classification of the third image, and judges the abnormality level of the pantograph in the secondary assessment period according to the assessment value and abnormality classification of the third image; when the abnormality level in the secondary assessment period reaches the system early warning standard, the abnormality handling module triggers the system abnormality early warning and the monitoring ends; when the abnormality level in the secondary assessment period reaches the system normal standard, the first judgment module monitors the pantograph with the normal state monitoring scheme; when the abnormality level in the secondary assessment period is between normal and early warning, the second judgment module continues to monitor the pantograph with the abnormality level assessment monitoring scheme; The monitoring method comprises the following steps: S1, set the monitoring time T i Get the i-th image U i , get U i The evaluation value Q i ; According to the basic evaluation value Q0 of the image; When Q i ≥Q0, the image U i If it is determined to be an abnormal image, the process goes to step S2; S2, monitor the pantograph with the abnormality evaluation monitoring scheme E1 during the abnormality evaluation period △T1, and set a total of m1 images U, U=[U1, U2, U3, ..., U m1 ], where the jth image is U j , get image U j The evaluation value Q j ; According to the basic evaluation value Q0, the evaluation interval threshold Q of the abnormality evaluation period is obtained max and Q min , where Q min ≤ln2Q0;Q max ≥2Q0-Q min ; When Q j ≥Q max When image U j It is judged to be a severe abnormality; When Q j ≤Q min When image U j Judged as normal; When Q min <Q j <Q max When image U j Judgement of mild abnormality; S3, among the m1 images acquired during the abnormality assessment period △T1, including c1 severely abnormal images and c2 slightly abnormal images, set the severe abnormality coefficient ω1 and the slightly abnormality coefficient ω2, where ω1>ω2; Obtain the abnormal distribution p1 = (c1 + c2) / m1 during the abnormality evaluation period; Among them, set the j1th severely abnormal image U j1 The evaluation value is Q j1 , the j2nd mildly abnormal image U j2 The evaluation value is Q j2 ; Set the evaluation threshold score W0; Obtain the evaluation score of the abnormality evaluation period △T1 ; When W1≥W0, the pantograph is judged to be severely abnormal, the system alarms and intervenes; When W1<W0, go to step S4; S4, monitor the pantograph with the secondary evaluation scheme during the secondary evaluation period △T2, and obtain m2 images, of which the zth image U z The evaluation value is Q z ; Among the m2 images, there are d1 severely abnormal images and d2 slightly abnormal images, and the abnormality distribution during the second evaluation period is obtained as p2 = (d1 + d2) / m2; Among them, set the z1th severely abnormal image U z1 The evaluation value is Q z1 , the z2nd mildly abnormal image U z2 The evaluation value is Q z2 ; Obtain assessment scores for the second assessment period ; When W2≥W0, the pantograph is judged to be severely abnormal, the system alarms and intervenes; When W2≤W1, the normal state monitoring scheme is restored; When W1<W2<W0, the secondary evaluation plan is released and monitoring is continued with the abnormality degree evaluation monitoring plan.

2. The monitoring method of the intelligent safety monitoring system of the railway pantograph network according to claim 1, characterized in that: It also includes an acquisition module, a processing module and an analysis module; wherein the acquisition module is used to acquire images of the pantograph during train operation; the processing module can process the acquired images to obtain key areas for judging the status of the pantograph; the analysis module can analyze the processed images to obtain an evaluation value representing the status of the pantograph.

3. The monitoring method of the intelligent safety monitoring system of the railway pantograph network according to claim 2, characterized in that: The processing module processes the collected image in the following steps: cropping, preprocessing, edge detection and edge extraction.

4. The monitoring method of the intelligent safety monitoring system of the railway pantograph network according to claim 2, characterized in that: The analysis module uses a deep learning model or a traditional machine vision algorithm to perform data analysis on the image processed by the processing module, thereby obtaining an evaluation value of the image.

Images