Online monitoring equipment for pantograph-catenary arc and detection system thereof

Abstract

The invention belongs to the technical field of railway electrical, and relates to online monitoring equipment for pantograph-catenary arcs and a detection system thereof, which comprises a portable case, an embedded industrial control all-in-one machine, a camera, a data storage card, a power supply module, an alarm device, a wireless communication module, a GPS (Global Positioning System) positioning module and an ardui no development board, the embedded industrial control all-in-one machine, the camera, the data storage card, the power module, the alarm device, the wireless communication module, the GPS positioning module and the ardui no development board are all installed in the portable case, accurate monitoring and early warning of pantograph-catenary arcs can be achieved under various complex environment backgrounds under the training of a neural network model, and meanwhile, after an RIO operator is introduced for parameter adjustment, accurate monitoring and early warning of the pantograph-catenary arcs can be achieved. The arc intensity can be subjected to grade measurement, early intervention suggestions can be provided for a railway operation and maintenance unit for fault early warning and operation and maintenance of the pantograph-catenary system in combination with a big data technology, actual reference values are provided, and the detection efficiency of the pantograph-catenary system is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of railway electrics, and specifically to an on-line monitoring device for pantograph-catenary arcs and its detection system. Background Art

[0002] The pantograph-catenary system is one of the key components of current electrified railways. The electric energy required by electric multiple units is obtained through continuous sliding contact between the pantograph on the roof and the overhead catenary. During the operation of the train, if affected by other uncontrollable factors such as poor static contact force or excessive load friction, it is very likely to cause arc discharge and wear of contact components, which will then lead to serious consequences such as affecting train power supply or interfering with train operation. Although railway operation and maintenance enterprises have a set of mature maintenance plans for pantograph wear, due to the inevitable and unpredictable characteristics of pantograph-catenary arcs, there is no good method to predict the service life of pantographs. Therefore, real-time status monitoring of key components such as pantographs and catenaries is an important technical means for the intelligent maintenance of electric multiple units, which is of great significance for train operation safety and economy, especially early fault diagnosis (including fault detection, location, degree assessment, etc.) has important reference value.

[0003] At present, domestic electric multiple units have adopted the high-speed railway power supply safety detection and monitoring system, namely the 6C system, which uses cameras to record and monitor the operation of the pantograph-catenary system, and analyzes the monitored videos and images. According to the investigation, it is found that at present, railway operation and maintenance enterprises still analyze video images by means of manual video recording playback, etc., and cannot qualitatively and quantitatively analyze pantograph-catenary arc information. Therefore, how to accurately and effectively extract pantograph-catenary arc information from the collected monitoring videos and scientifically and reliably analyze the data is an important prerequisite for railway infrastructure maintenance and fault diagnosis.

[0004] Therefore, developing an intelligent safety protection device combined with an advanced device system is crucial for improving the safety of power grid maintenance.

[0005] In view of the above technical deficiencies, a solution for an on-line monitoring device for pantograph-catenary arcs and its detection system is now proposed. Summary of the Invention

[0006] To solve the above problems, the present invention provides the following technical solutions:

[0007] An on-line monitoring device for pantograph-catenary arcs, including a portable chassis, an embedded industrial computer, a camera, a data memory card, a power module, an alarm device, a wireless communication module, a GPS positioning module, and an Arduino development board. The embedded industrial computer, the camera, the data memory card, the power module, the alarm device, the wireless communication module, the GPS positioning module, and the Arduino development board are all installed in the portable chassis.

[0008] Furthermore, the embedded industrial control all-in-one computer is developed based on the Labview development environment, and uses the built-in VDM and VDS functions to realize functions such as camera image data acquisition, image feature analysis, RIO operator parameter adjustment, neural network model construction, and other functions.

[0009] Furthermore, the camera is used to capture image information of the pantograph-catenary contact area in real time.

[0010] Furthermore, the data storage card is used to store the image data captured by the camera and the data processed by the industrial control all-in-one computer.

[0011] Furthermore, the wireless communication module is used to transmit the monitoring data to the remote monitoring center in real time.

[0012] Furthermore, the GPS positioning module is used to determine the geographical location information of the pantograph-catenary arcing.

[0013] Furthermore, the arduino development board is used to control and coordinate the work of each module.

[0014] Furthermore, the remote monitoring center receives real-time data from the monitoring device and performs data analysis and processing. The remote monitoring center realizes the real-time processing of data and the later maintenance of the pantograph-catenary arc through the neural network training model, RIO operator parameter adjustment, arc intensity evaluation and arc level division, and fault warning and intervention suggestions. The image information collected by the camera is defined as Q z , and the collected information is processed by the embedded industrial control all-in-one computer. The processed information is defined as Q 标准 , and this information is continuously trained through the neural network model. The specific neural network model is as follows:

[0015]

[0016] where N is the number of samples, y i is the true label, is the model prediction value, and θ is the parameter of the neural network;

[0017] At the same time, the optimization of the neural network uses the gradient descent algorithm, and the update rule is as follows:

[0018]

[0019] where η is the learning rate, is the gradient of the loss function with respect to the model parameters;

[0020] The RIO operator parameter adjustment optimizes the input features in the neural network to improve the accuracy of arc monitoring. The optimization process depends on the objective function J and the feature selection strategy, specifically as follows:

[0021]

[0022] where x ij represents the j-th feature of the i-th sample, ω j represents the weight of the j-th feature, b is the bias term, N is the number of samples, and M is the feature dimension;

[0023] Adjust the parameters of the RIO operator to find the best combination of input features so that the arc intensity and type can be accurately predicted;

[0024] Regarding the arc intensity evaluation and arc level classification, the arc intensity evaluation is based on the brightness and area of the arc region. Considering the diversity of images, it can be calculated by weighted average. The formula is as follows:

[0025]

[0026] where A is the number of pixels in the arc region, W and H are the width and height of the image respectively, I arc (x, y) is the pixel brightness value of the arc region, and ω(x, y) is the weighting factor of the position, which is used to enhance the influence of important regions.

[0027] Furthermore, regarding the arc level classification, the arc intensity S is used to classify the arcs. The arc intensity is divided into three levels: low, medium, and high. The classification formula is as follows:

[0028]

[0029] where S low , S high are the thresholds for low intensity and high intensity respectively;

[0030] Regarding the fault warning and intervention suggestions, by combining big data applications, the fault warning is based on the arc intensity and occurrence probability. When the arc intensity exceeds the set threshold and the occurrence probability exceeds the warning threshold, the system will generate fault intervention suggestions. The specific warning formula is as follows:

[0031]

[0032] where S th is the warning threshold of the arc intensity, P th is the warning threshold of the arc occurrence probability, P arc is the arc occurrence probability. At the same time, based on big data analysis, the generated intervention suggestions can optimize the maintenance decision through the following formula:

[0033]

[0034] where is the weight, is a feature related to the fault, and Risk(i) is the risk assessment value related to the i-th feature;

[0035] Regarding the real-time monitoring and feedback, through real-time data collection and analysis, continuously monitor the arc state of the electric multiple unit, and conduct dynamic evaluation. According to the real-time data and feedback, the system will dynamically adjust the prediction parameters of the neural network and the strategy of fault warning to ensure efficient response to different situations.

[0036] Furthermore, the specific steps are as follows:

[0037] Step 1: Use a camera to monitor the arcing situation of the electric multiple unit in real time, collect relevant data of the arc, preprocess the collected arc images, and extract the arc area from the background for subsequent analysis;

[0038] Step 2: Select a suitable neural network architecture (such as the convolutional neural network CNN) according to the requirements to process the arc images, label the collected arc images into different categories (such as normal arc, minor arc, severe arc, etc.), select a suitable loss function, and optimize the model through the backpropagation algorithm and the gradient descent method;

[0039] Step 3: Analyze the arc area using the features output by the middle layer of the neural network to extract effective information;

[0040] Step 4: Evaluate the arc intensity value by calculating the brightness and area of the arc area, etc. The arc intensity formula S can be used to quantify the arc intensity. According to the arc intensity value S, adopt a preset grade standard to divide the arc into low, medium, and high intensity grades;

[0041] Step 5: Use historical data and arc intensity, and use the big data analysis model to calculate the probability of arc occurrence. When the arc intensity exceeds the preset threshold and the occurrence probability reaches the warning threshold, the system automatically generates a fault warning message. Based on information such as the arc intensity and occurrence probability, the system generates targeted intervention suggestions to help the operation and maintenance personnel take measures in time;

[0042] Step 6: Through real-time data collection and analysis, continuously monitor the arc state of the electric multiple unit, and conduct dynamic evaluation. According to the real-time data and feedback, the prediction parameters of the neural network and the fault warning strategy will be dynamically adjusted to ensure efficient response to different situations;

[0043] Step 7: Integrate the image acquisition device with the computing platform to ensure that data can be transmitted to the remote monitoring center in time, design a user-friendly interface for the operation and maintenance personnel, and display the real-time state of the arc, the intensity evaluation result, the warning information, and the intervention suggestions.

[0044] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0045] 1. In the on-line monitoring device and its detection system for pantograph-catenary arcs of the present invention, under the training of the neural network model, accurate monitoring and early warning of pantograph-catenary arcs can be realized in various complex environmental backgrounds. After introducing the RIO operator for parameter adjustment, the arc intensity can be graded. Combining with big data applications, early fault repair intervention suggestions can be provided for railway operation and maintenance units for fault early warning and operation and maintenance repair of the pantograph-catenary system, providing practical reference value.

[0046] 2. In the on-line monitoring device and its detection system for pantograph-catenary arcs of the present invention, serial communication with LabVIEW is realized through an Arduino development board, a wireless communication module, a GPS positioning module, a data memory card, etc., and the detection alarm data can be stored in real time, uploaded to the data cloud, and the geographical location of the fault point can be determined, which is convenient for operation and maintenance units to call the data for later review and repair of the pantograph-catenary system.

[0047] 3. In the on-line monitoring device and its detection system for pantograph-catenary arcs of the present invention, in order to increase the portability of the device and reduce data cache redundancy, three data acquisition and detection modes are added, which can be adjusted according to needs, and a camera is used to collect the data set. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings;

[0049] Figure 1 is a flowchart of the steps of the detection system for pantograph-catenary arcs of the present invention;

[0050] Figure 2 is a schematic diagram of the overall structure of the on-line monitoring device for pantograph-catenary arcs of the present invention.

[0051] Reference numerals: 1, portable chassis; 2, embedded industrial personal computer; 3, camera; 4, data memory card; 5, controller main body; 6, alarm device; 7, wireless communication module; 8, GPS positioning module; 9, Arduino development board. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0053] As Figure 1 - Figure 2As shown in the figure, the on-line monitoring device for pantograph-catenary arc includes a portable chassis 1, an embedded industrial control computer 2, a camera 3, a data memory card 4, a power module 5, an alarm device 6, a wireless communication module 7, a GPS positioning module 8, and an Arduino development board 9. The embedded industrial control computer, the camera, the data memory card, the power module, the alarm device, the wireless communication module, the GPS positioning module, and the Arduino development board are all installed in the portable chassis.

[0054] Based on the Labview development environment, the embedded industrial control computer uses the built-in VDM and VDS functions to realize functions such as camera image data acquisition, image feature analysis, RIO operator parameter adjustment, construction of a neural network model, and other functions.

[0055] The camera 3 is used to capture the image information of the pantograph-catenary contact area in real time.

[0056] The data memory card 4 is used to store the image data captured by the camera and the data processed by the industrial control computer.

[0057] The wireless communication module 7 is used to transmit the monitoring data to the remote monitoring center in real time.

[0058] The GPS positioning module is used to determine the geographical location information of the monitoring device.

[0059] The Arduino development board 9 is used to control and coordinate the work of each module.

[0060] The remote monitoring center receives the real-time data from the monitoring device and conducts data analysis and processing. The remote monitoring center realizes the real-time processing of data and the later maintenance of the pantograph-catenary arc through the neural network training model, RIO operator parameter adjustment, arc intensity evaluation and arc level division, and fault warning and intervention suggestions. The image information collected by the camera 3 is defined as Q z and the collected information is processed through the embedded industrial control computer. The processed information is defined as Q 标准 and this information is continuously trained through the neural network model. The specific neural network model is as follows:

[0061]

[0062] where N is the number of samples, y i is the true label, is the model prediction value, and θ is the parameter of the neural network;

[0063] At the same time, the optimization of the neural network uses the gradient descent algorithm, and the update rule is as follows:

[0064]

[0065] where η is the learning rate, is the gradient of the loss function with respect to the model parameters;

[0066] The parameter tuning of the RIO operator optimizes the input features in the neural network, improves the accuracy of arc monitoring, and the optimization process depends on the objective function J and the feature selection strategy, as follows:

[0067]

[0068] where x ij , represents the j-th feature of the i-th sample, ω j represents the weight of the j-th feature, b is the bias term, N is the number of samples, and M is the feature dimension;

[0069] Adjust the parameters of the RIO operator such as the learning rate, regularization coefficient, etc. to find the best combination of input features so that the arc intensity and type can be accurately predicted;

[0070] Use image processing algorithms such as convolutional neural networks to analyze arc images in real time, output the arc intensity, and based on the real-time collected arc data, the system dynamically adjusts the input features to improve the accuracy of arc detection.

[0071] Arc intensity evaluation and arc level classification. The arc intensity evaluation is based on the brightness and area of the arc region. Considering the diversity of images, it can be calculated by weighted average. The formula is as follows:

[0072]

[0073] where A is the number of pixels in the arc region, W and H are the width and height of the image respectively, and I arc (x, y) is the pixel brightness value of the arc region, and ω(x, y) is the weighting factor of the position, which is used to enhance the influence of important regions.

[0074] Real-time image processing. Obtain arc images through cameras or other sensors, analyze the brightness values of each pixel point, and perform weighted calculations according to the set weights. Use image processing algorithms such as convolutional neural networks to analyze arc images in real time and output the arc intensity.

[0075] Arc level classification. The arc intensity S is used to classify the arc. The arc intensity is set to be divided into three levels: low, medium, and high. The classification formula is as follows:

[0076]

[0077] where S low 、S high are the thresholds for low intensity and high intensity respectively;

[0078] Fault warning and intervention suggestions are generated by combining big data applications. Fault warning is based on arc intensity and occurrence probability. When the arc intensity exceeds the set threshold and the occurrence probability exceeds the warning threshold, the system will generate fault intervention suggestions. The specific warning formula is as follows:

[0079]

[0080] Where S th is the warning threshold of arc intensity, and P th is the warning threshold of arc occurrence probability. P arc is the arc occurrence probability. At the same time, based on big data analysis, the generated intervention suggestions can optimize the maintenance decision through the following formula:

[0081]

[0082] Where is the weight, is the feature related to the fault, and Risk(i) is the risk assessment value related to the i-th feature;

[0083] Using real-time sensor data and big data analysis to predict the arc occurrence probability, the system can process arc data in real time and automatically generate warning notifications according to the set warning thresholds, providing maintenance suggestions. When the arc intensity is greater than the set threshold and the occurrence probability exceeds the warning threshold, the system will send a warning.

[0084] Real-time monitoring and feedback. Through real-time data collection and analysis, continuously monitor the arc state of the electric multiple unit and conduct dynamic evaluation. According to real-time data and feedback, the system will dynamically adjust the prediction parameters of the neural network and the fault warning strategy to ensure efficient response to different situations.

[0085] The specific steps are as follows:

[0086] Step 1: Use a camera to monitor the arc ignition situation of the electric multiple unit in real time, collect relevant arc data, and perform preprocessing such as denoising, enhancement, and segmentation on the collected arc images to extract the arc area from the background for subsequent analysis;

[0087] Step 2: Select a suitable neural network architecture such as the convolutional neural network CNN according to the needs to process the arc images. Label the collected arc images into different categories such as normal arc, minor arc, severe arc, etc., select a suitable loss function, and optimize the model through the backpropagation algorithm and the gradient descent method;

[0088] Step 3: Analyze the arc region using the features output by the middle layer of the neural network to extract effective information. Step 4: Evaluate the arc intensity value by calculating the brightness and area of the arc region. The arc intensity formula S can be used to quantify the arc intensity. According to the arc intensity value S, the arc is classified into low, medium, and high intensity levels using a preset grading standard.

[0089] Step 5: Use historical data and arc intensity to calculate the probability of arc occurrence using a big data analysis model. When the arc intensity exceeds the preset threshold and the occurrence probability reaches the warning threshold, the system automatically generates a fault warning message. Based on information such as the arc intensity and occurrence probability, the system generates targeted intervention suggestions to help maintenance personnel take timely measures.

[0090] Step 6: The system continuously monitors the arc state of the electric multiple unit through real-time data collection and analysis and conducts dynamic evaluation. Based on real-time data and feedback, the system dynamically adjusts the prediction parameters of the neural network and the fault warning strategy to ensure efficient response to different situations.

[0091] Step 7: Integrate the image acquisition device with the computing platform to ensure that data can be transmitted to the remote monitoring center in a timely manner. Design a user-friendly interface for maintenance personnel to display the real-time state of the arc, the intensity evaluation result, the warning message, and the intervention suggestion.

[0092] Working principle: First, an embedded industrial control all-in-one computer is used to control the camera to accurately capture the image information of the pantograph-catenary contact area in real time, and the captured information is transmitted into the embedded industrial control all-in-one computer. After receiving the captured information, the embedded industrial control all-in-one computer processes the image data captured by the camera through the built-in Vision Development Module and Vision Acquisition Software functions. First, preprocess the image, including denoising and enhancement, to improve the image quality. Then, use the image feature analysis algorithm to extract the key features in the image, such as the shape and color of the arc. Next, use functions such as RIO operator parameter adjustment and neural network call to further analyze and process the extracted features, so as to accurately judge the generation of pantograph-catenary arc. When the pantograph-catenary arc is detected, the alarm device immediately gives an audible and visual alarm to remind the operation and maintenance personnel to pay attention. At the same time, the wireless communication module transmits the monitoring data (including the geographical location of the fault point) to the remote monitoring center in real time. The remote monitoring center further analyzes and processes the received data, including data cleaning, feature extraction, and model training, to achieve comprehensive monitoring and early warning of the pantograph-catenary arc. According to the analysis results, the remote monitoring center can provide suggestions for early fault repair intervention to the operation and maintenance unit to help the operation and maintenance personnel deal with potential faults in time. In addition, the remote monitoring center also has the function of uploading data to the cloud. The operation and maintenance unit can access the stored monitoring data through the cloud platform for data review and analysis to better understand the operation status of the pantograph-catenary system and improve the operation and maintenance efficiency.

[0093] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to the specific embodiments. Obviously, many modifications and changes can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the present invention, so that those skilled in the art in the relevant technical field can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. An on-line monitoring device for pantograph-catenary arcs, characterized in that It includes a portable chassis (1), an embedded industrial computer (2), a camera (3), a data memory card (4), a power module (5), an alarm device (6), a wireless communication module (7), a GPS positioning module (8), and an Arduino development board (9). The embedded industrial computer, camera, data memory card, power module, alarm device, wireless communication module, GPS positioning module, and Arduino development board are all installed inside the portable chassis.

2. The on-line monitoring device for pantograph-catenary arcs according to claim 1, characterized in that, The embedded industrial computer is based on the Labview development environment and uses the built-in VDM and VDS functions to achieve functions such as camera image data acquisition, image feature analysis, RIO operator parameter adjustment, neural network model construction, and other functions.

3. The on-line monitoring device for pantograph-catenary arcs according to claim 1, characterized in that, The camera (3) is used to capture the image information of the pantograph-catenary contact area in real time.

4. The on-line monitoring device for pantograph-catenary arc according to claim 3, characterized in that, The data memory card (4) is used to store the image data captured by the camera and the data processed by the industrial computer.

5. The on-line monitoring device for pantograph-catenary arc according to claim 1, characterized in that, The wireless communication module (7) is used to transmit the monitoring data to the remote monitoring center in real time.

6. The on-line monitoring device for pantograph-catenary arcs according to claim 1, characterized in that The GPS positioning module is used to determine the geographical location information of the pantograph-catenary arcing.

7. The on-line monitoring device for pantograph-catenary arcs according to claim 1, characterized in that, The Arduino development board (9) is used to control and coordinate the work of each module.

8. A detection system for pantograph-catenary arcs, which uses the on-line monitoring device for pantograph-catenary arcs and the remote monitoring center described in any one of the above claims 1-7, is characterized in that, The remote monitoring center receives real-time data from the monitoring devices and conducts data analysis and processing. The remote monitoring center realizes the real-time processing of data and the later maintenance of pantograph-catenary arcs through a neural network training model, RIO operator parameter tuning, arc intensity evaluation and arc level classification, and fault warning and intervention suggestions. The image information collected by the camera (3) is defined as Q z , and the collected information is processed through an embedded industrial computer. The information after processing is defined as Q 标准 , and the information is continuously trained through a neural network model. The specific neural network model is as follows: where N is the number of samples, y i is the true label, is the model prediction value, and θ are the parameters of the neural network At the same time, the optimization of the neural network uses the gradient descent algorithm, and the update rule is as follows: θ t+1 = θ t - η·▽ θ L(θ t ) where η is the learning rate, and ▽ θ L(θ t ) is the gradient of the loss function with respect to the model parameters; The RIO operator parameter adjustment optimizes the input features in the neural network to improve the accuracy of arc monitoring. The optimization process depends on the objective function J and the feature selection strategy, specifically as follows: where x ij , represents the j-th feature of the i-th sample, ω j represents the weight of the j-th feature, b is the bias term, N is the number of samples, and M is the feature dimension; Adjust the parameters of the RIO operator (such as the learning rate, regularization coefficient, etc.) to find the best combination of input features so that the arc intensity and type can be accurately predicted; The arc intensity evaluation and arc level division. The arc intensity evaluation is based on the brightness and area of the arc region. Considering the diversity of images, it can be calculated by the weighted average method. The formula is as follows: Where A is the number of pixels in the arc region, W and H are the width and height of the image respectively, and I arc (x, y) is the pixel brightness value of the arc region, and ω(x, y) is the weighting factor of the position, which is used to enhance the influence of the important region.

9. The detection system for pantograph-catenary arcs according to claim 1, characterized in that, For the arc level division, the arc intensity S is used to divide the arc into three levels: low, medium, and high. The division formula is as follows: where S low and S high are the low-intensity and high-intensity thresholds, respectively; The fault warning and intervention suggestions are through the combination of big data applications. The fault warning is based on the arc intensity and occurrence probability. When the arc intensity exceeds the set threshold and the occurrence probability exceeds the warning threshold, the system will generate fault intervention suggestions. The warning formula is as follows: where S th is the warning threshold of the arc intensity, P th is the warning threshold of the arc occurrence probability, P arc is the arc occurrence probability. Meanwhile, according to big data analysis, the generated intervention suggestions can optimize the maintenance decision through the following formula: wherein is the weight, is the feature related to the fault, and Risk(i) is the risk assessment value related to the i-th feature; The real-time monitoring and feedback, through real-time data collection and analysis, continuously monitors the arc state of the electric multiple unit and conducts dynamic evaluation. According to the real-time data and feedback, the system will dynamically adjust the prediction parameters of the neural network and the fault warning strategy to ensure efficient response to different situations.

10. The detection system for pantograph-catenary arc according to claim 1, characterized in that, The specific steps are as follows: Step 1: Use the camera to monitor the arcing situation of the electric multiple unit in real time, collect the relevant data of the arc, preprocess the collected arc images, and extract the arc region from the background; Step 2: Select a neural network according to the needs to process the arc images, label the collected arc images into different categories, select an appropriate loss function, and optimize the model through the backpropagation algorithm and the gradient descent method; Step 3: Analyze the arc region using the features output by the middle layer of the neural network to extract effective information; Step 4: Evaluate the arc intensity value by calculating the brightness and area of the arc region. Use the arc intensity formula S to quantify the arc intensity. According to the arc intensity value S, adopt a preset grading standard to divide the arc into low, medium, and high intensity levels; Step 5: Use historical data and arc intensity, and apply a big data analysis model to calculate the probability of arc occurrence. When the arc intensity exceeds the preset threshold and the occurrence probability reaches the warning threshold, the system automatically generates a fault warning message. Based on the arc intensity and occurrence probability information, the system generates targeted intervention suggestions; Step 6: Through real-time data collection and analysis, continuously monitor the arc state of the electric multiple unit and conduct dynamic evaluation. According to the real-time data and feedback, dynamically adjust the prediction parameters of the neural network and the fault warning strategy; Step 7: Integrate the image acquisition device with the computing platform to display the real-time state of the arc, the intensity evaluation result, the warning message, and the intervention suggestion.

Images