Intelligent detection method and system for pantograph-catenary wear based on machine vision

Abstract

The present application relates to the technical field of computer vision analysis, and specifically includes a pantograph-catenary wear intelligent detection method and system based on machine vision, which comprises: obtaining pantograph state and image information in the pantograph-catenary contact area, inverting the fatigue crack propagation path under pantograph lifting mechanism control parameters, extracting space-time features through three-dimensional convolution, combining environmental compensation and image information to fuse multi-scale features, generating dynamic wear feature tensors for wear deviation control, solving the technical problems that the fatigue crack propagation path is interfered by various factors and the pantograph and catenary wear detection precision is insufficient, realizing the fatigue crack propagation path of the pantograph lifting mechanism and pantograph lowering mechanism, comprehensively evaluating the wear condition of the pantograph and catenary, remotely detecting and maintaining the pantograph and catenary in real time, timely discovering and processing potential wear hazards, avoiding accidents, and improving the technical effects of detection precision and efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer vision analysis, and particularly relates to a pantograph-catenary wear intelligent detection method and system based on machine vision. BACKGROUND

[0002] The pantograph-catenary is used to deliver electric energy of a traction catenary to an electric locomotive. Since the working environment is harsh and the pantograph-catenary often bears strong electrical and mechanical effects, various diseases are prone to occur. The wear not only affects the performance of the pantograph and the catenary, but also causes power supply failure and affects the normal operation of the electric locomotive.

[0003] With the rapid development of computer technology and machine vision, detection gradually develops in the direction of online detection and real-time detection. Remote monitoring of the pantograph is performed through an image device, and networking is performed with a subway maintenance center remotely, so that subway maintenance personnel can timely and accurately find faults of the pantograph. However, in actual application, there are still problems such as poor adaptability to environmental factors and insufficient recognition ability for complex diseases.

[0004] In summary, in the prior art, the extension path of the fatigue crack is disturbed by various factors, and the wear detection precision of the pantograph and the catenary is insufficient. SUMMARY

[0005] The present application provides a pantograph-catenary wear intelligent detection system based on machine vision, which aims to solve the technical problems of the prior art that the extension path of the fatigue crack is disturbed by various factors and the wear detection precision of the pantograph and the catenary is insufficient.

[0006] In view of the above problems, the technical scheme of the present application is:

[0007] In an aspect of the present application, a machine vision-based pantograph-catenary wear intelligent detection method is provided, wherein the method comprises: obtaining the working state and surface image information of the pantograph in a pantograph-catenary contact area between the catenary wire and the slide plate; performing finite element simulation inversion on the lifting pantograph mechanism control parameters corresponding to the upper frame and the lower arm rod and the lowering pantograph mechanism control parameters corresponding to the upper frame and the lower arm rod, respectively, to obtain a first fatigue crack propagation path set under a plurality of control parameter combinations and a second fatigue crack propagation path set under a plurality of control parameter combinations; performing three-dimensional convolution processing on the first fatigue crack propagation path set and the second fatigue crack propagation path set to extract the spatiotemporal correlation features; adding environmental compensation data to correct the perception deviation based on the working state and surface image information of the pantograph, and combining the first fatigue crack propagation path set and the second fatigue crack propagation path set to perform multi-scale feature fusion to generate a dynamic wear feature tensor under the spatiotemporal correlation features; and performing wear offset control based on the dynamic wear feature tensor in combination with the lifting pantograph mechanism control parameters and the lowering pantograph mechanism control parameters.

[0008] In another aspect of the present application, a machine vision-based pantograph-catenary wear intelligent detection system is provided, wherein the system comprises: an information acquisition module for obtaining the working state and surface image information of the pantograph in a pantograph-catenary contact area between the catenary wire and the slide plate; a simulation module for performing finite element simulation inversion on the lifting pantograph mechanism control parameters corresponding to the upper frame and the lower arm rod and the lowering pantograph mechanism control parameters corresponding to the upper frame and the lower arm rod, respectively, to obtain a first fatigue crack propagation path set under a plurality of control parameter combinations and a second fatigue crack propagation path set under a plurality of control parameter combinations; a convolution processing module for performing three-dimensional convolution processing on the first fatigue crack propagation path set and the second fatigue crack propagation path set to extract the spatiotemporal correlation features; a feature fusion module for adding environmental compensation data to correct the perception deviation based on the working state and surface image information of the pantograph, and combining the first fatigue crack propagation path set and the second fatigue crack propagation path set to perform multi-scale feature fusion to generate a dynamic wear feature tensor under the spatiotemporal correlation features; and an offset control module for performing wear offset control based on the dynamic wear feature tensor in combination with the lifting pantograph mechanism control parameters and the lowering pantograph mechanism control parameters.

[0009] In summary, one or more technical solutions provided in the present application achieve the following technical effects: the fatigue crack propagation paths of the lifting pantograph mechanism and the lowering pantograph mechanism are inverted, the wear of the pantograph-catenary is comprehensively evaluated, the real-time detection and maintenance decision of the remote wear of the pantograph and the catenary are performed, potential wear hazards are found and handled in a timely manner to avoid accidents, and the detection precision and efficiency are improved. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1A flowchart of an intelligent detection method for pantograph-catenary wear based on machine vision is provided for the present application.

[0011] Figure 2 A structural diagram of an intelligent detection system for pantograph-catenary wear based on machine vision is provided for the present application.

[0012] Legend: information acquisition module M100, simulation module M200, convolution processing module M300, feature fusion module M400, offset control module M500. DETAILED DESCRIPTION

[0013] EMBODIMENT

[0014] The present application will be specifically described below in conjunction with the accompanying drawings, as shown in the drawings, the present application provides an intelligent detection method for pantograph-catenary wear based on machine vision, wherein the method comprises: Figure 1

[0015] S1: obtaining the working state and surface image information of the pantograph in the pantograph-catenary contact area between the catenary conductor and the slide plate; S2: performing finite element simulation inversion on the control parameters of the pantograph corresponding to the upper frame and the lower arm, and the control parameters of the pantograph corresponding to the upper frame and the lower arm, respectively, to obtain a first fatigue crack propagation path set under a plurality of control parameter combinations and a second fatigue crack propagation path set under a plurality of control parameter combinations.

[0016] Specifically, obtaining the working state and surface image information of the pantograph in the pantograph-catenary contact area means using a high-resolution camera and an imaging system to real-time capture the working state and surface image of the pantograph, and pre-processing the obtained image through image processing algorithms, including denoising, enhancing contrast, image segmentation, etc., to highlight the wear area and wear degree; finite element simulation inversion is a numerical simulation method based on finite element analysis, which discretizes the actual structure, establishes a finite element model, applies corresponding loads and boundary conditions, solves the response of the structure, and further obtains the mechanical properties and damage evolution of the structure under different working conditions; fatigue crack propagation path refers to the trajectory of the internal crack of the material gradually expanding under cyclic loading, and through the analysis of the crack propagation path, the fatigue life and failure mode of the material can be predicted.

[0017] ​A high-resolution camera and imaging system are installed in the catenary wire and the pantograph contact area between the slide plate to obtain real-time pantograph working state and surface image information; according to the structure and working principle of the pantograph, the control parameters of the pantograph lifting mechanism and the pantograph lowering mechanism are determined, including air pressure value, pantograph lifting speed, pantograph lowering angle, etc.; based on these control parameters, finite element simulation inversion is carried out respectively, different control parameter combinations are set, and the stress condition and crack propagation process of the pantograph under various working conditions are simulated; through finite element analysis software (such as ABAQUS, MARC, etc.), the finite element model of the pantograph is established, the corresponding load and boundary conditions are applied, and the first fatigue crack propagation path set and the second fatigue crack propagation path set under multiple control parameter combinations are obtained.

[0018] By obtaining high-definition image information of the pantograph, a rich data basis is provided for subsequent wear feature extraction and analysis. The wear marks and cracks on the surface of the pantograph contained in the image intuitively reflect the wear condition of the pantograph; by using finite element simulation inversion technology, the fatigue crack propagation path of the pantograph lifting and lowering mechanism is predicted, which can master the fatigue damage of the pantograph under different working conditions in advance and provide support for targeted optimization.

[0019] S3: Perform three-dimensional convolution processing on the first fatigue crack propagation path set and the second fatigue crack propagation path set to extract spatiotemporal correlation features; S4: Based on the pantograph working state and surface image information, add environmental compensation data to correct perception deviation, and combine the first fatigue crack propagation path set and the second fatigue crack propagation path set to perform multi-scale feature fusion to generate a dynamic wear feature tensor under spatiotemporal correlation features; S5: Based on the dynamic wear feature tensor, combine the pantograph lifting mechanism control parameters and the pantograph lowering mechanism control parameters to perform wear offset control.

[0020] Specifically, the three-dimensional convolution processing of the first fatigue crack propagation path set and the second fatigue crack propagation path set refers to feature extraction and analysis of these path sets by using a three-dimensional convolutional neural network (3D CNN); three-dimensional convolution can perform convolution operations in space and time dimensions simultaneously, capture motion information and spatial features between consecutive frames, and thus extract spatiotemporal correlation features of crack propagation; adding environmental compensation data to correct perception bias refers to compensating the acquired image information by introducing environmental parameters (such as temperature, humidity, and illumination) to reduce the influence of environmental factors on the detection results; multi-scale feature fusion refers to integrating features of different scales and levels to more comprehensively describe the features of crack propagation; generating a dynamic wear feature tensor refers to representing the fused features in the form of a tensor for subsequent analysis and processing; and wear offset control is achieved by actively adjusting the pantograph raising mechanism control parameters and the pantograph lowering mechanism control parameters to maintain the wear position and degree of the pantograph key components in a controllable state and avoid local excessive wear.

[0021] The three-dimensional convolution processing of the first fatigue crack propagation path set and the second fatigue crack propagation path set regards these path sets as continuous three-dimensional data blocks, uses a three-dimensional convolution kernel to perform convolution operations in space and time dimensions, extracts gradient features, multi-band features, and other spatiotemporal correlation features of crack propagation; based on the pantograph working state and surface image information, environmental compensation data is added to correct perception bias, environmental parameters such as temperature, humidity, and illumination intensity are collected, an environmental compensation model is established, image information is adjusted, and detection errors caused by environmental factors are reduced. Specifically, in an outdoor detection system, illumination changes can affect image quality, and environmental compensation can improve the stability and reliability of images.

[0022] Multi-scale feature fusion is combined with the first fatigue crack propagation path set and the second fatigue crack propagation path set to integrate crack propagation features at different scales and form a more comprehensive feature description. Specifically, the concat operation in the early fusion method can concatenate different features to increase the dimensionality of the features; a dynamic wear feature tensor under spatiotemporal correlation features is generated to represent the fused features in the form of a tensor, and the dimensions of the tensor can include time, space, crack depth, etc., facilitating subsequent analysis and processing; by analyzing the wear features in the tensor and combining the current pantograph raising mechanism control parameters and pantograph lowering mechanism control parameters, the future wear offset trend of the pantograph can be predicted to change the pantograph contact force distribution and thus suppress wear offset.

[0023] Preferably, by three-dimensional convolution processing and spatio-temporal feature extraction, the dynamic process of crack propagation and the spatio-temporal variation characteristics can be captured, providing more accurate basis for wear evaluation. On the one hand, the addition of environmental compensation data corrects the perception deviation, improves the stability and reliability of the detection results, and enables the system to better adapt to complex and variable actual environments. Multi-scale feature fusion further integrates feature information of different scales, making the description of wear characteristics more comprehensive and detailed. The generated dynamic wear feature tensor provides rich data support for subsequent maintenance decisions. On the other hand, the configured pantograph and catenary maintenance control instruction set directly converts the detection results into specific maintenance action guidance, achieving precise maintenance of the pantograph and catenary, improving the efficiency and pertinence of maintenance work, and ensuring the safety and reliability of rail transit operation.

[0024] Further, based on the dynamic wear feature tensor, the wear offset control is performed in combination with the pantograph lifting mechanism control parameters and the pantograph lowering mechanism control parameters. The method of the application comprises:

[0025] Based on the dynamic wear feature tensor, the wear abnormal area and the wear depth evolution index are determined. According to the wear abnormal area and the wear depth evolution index, the feasible region of the pantograph lifting mechanism control parameters and the pantograph lowering mechanism control parameters is updated, and the feasible region is mapped to the parameter combination with the minimum wear offset obtained by searching.

[0026] Specifically, the wear abnormal area refers to the area where the wear degree exceeds the normal range by analyzing the data in the dynamic wear feature tensor. The wear depth evolution index is obtained by quantifying the change trend of the wear depth with time in the wear abnormal area. The feasible region refers to the reasonable range set of the pantograph lifting mechanism control parameters and the pantograph lowering mechanism control parameters in terms of value. The parameter combination in this range is considered to be able to meet the system operation requirements and comply with the wear offset control target.

[0027] By analyzing the dynamic wear feature tensor, the area with abnormal wear degree is identified by setting a wear threshold or using a clustering algorithm. For example, in tool wear detection, the wear point is determined by a 3D model. The wear depth evolution index, which can be the wear depth change rate with time or other indicators reflecting the wear development trend, is calculated. For example, the wear coefficient is estimated by the Archard model, and the wear depth is calculated based on the wear volume and sliding distance.

[0028] According to the position, size and wear depth evolution index of the abnormal wear area, combined with finite element simulation data, historical experience data and the like, the feasible region of the control parameters of the pantograph lifting mechanism and the control parameters of the pantograph lowering mechanism is redefined. Specifically, if a certain abnormal wear area shows that local wear is too fast, the range of the value of the pantograph lifting pressure parameter is reduced to avoid aggravating wear due to excessive pressure. Based on the wear state information obtained through the previous detection and analysis, the range for the subsequent optimization search of the control parameters is determined, and the blindness and calculation amount of the parameter search are effectively reduced. Reasonable definition of the feasible region can improve the efficiency of the control parameter optimization, enable the control parameter combination that minimizes the wear deviation to be found more quickly and accurately, and thus effectively control the wear deviation of the pantograph and catenary, prolong the service life of the pantograph and catenary, and ensure the stable operation of the rail transit power supply system.

[0029] The wear abnormal area and the wear depth evolution index are determined.

[0030] Further, for the wear maintenance decision of the pantograph and the catenary, the method of the application further includes:

[0031] In the updated feasible region, iterative search is performed with minimization of the wear deviation as a reward function; and in the iterative search process, the smoothness of the pantograph lifting mechanism control parameters and the smoothness of the pantograph lowering mechanism control parameters are used as penalty terms.

[0032] Specifically, the iterative search refers to a process of gradually approaching the optimal solution by repeatedly trying different combinations of the pantograph lifting mechanism control parameters and the pantograph lowering mechanism control parameters in the updated feasible region; the reward function refers to a mathematical function used to evaluate the advantages and disadvantages of each search result, and here the minimization of the wear deviation is used as the reward function, that is, the smaller the wear deviation, the higher the “reward” obtained, which means that the parameter combination is better; the penalty term is a mechanism for “penalizing” the search results that do not meet the expectations, and the smoothness of the pantograph lifting mechanism control parameters and the smoothness of the pantograph lowering mechanism control parameters are used as the penalty terms to avoid drastic fluctuations in the control parameters during the adjustment process.

[0033] In the updated and determined feasible region, an intelligent optimization algorithm (such as a genetic algorithm or a particle swarm optimization algorithm) is used to randomly or according to a certain strategy select an initial control parameter combination, which is substituted into the calculation model of the reward function and the penalty term. The minimization of the wear deviation is used as the reward function, and a pantograph and catenary wear prediction model (such as a model based on finite element analysis or machine learning) is established to obtain the wear deviation under the parameter combination, and the smaller the wear deviation, the higher the reward obtained.

[0034] Meanwhile, the change range of the pantograph-raising mechanism control parameter and the pantograph-lowering mechanism control parameter between adjacent iterations is calculated, and the greater the change range, the higher the value of the penalty term. Through continuous iteration, the parameter combination direction of the next search is adjusted according to the reward and penalty obtained each time, and gradually searches in the direction of maximizing the reward (minimizing the wear offset) and minimizing the penalty (high parameter smoothness).

[0035] By introducing the reward function and the penalty term, when searching for the optimal control parameter combination, not only the reduction of the wear offset is considered, but also the stability of the parameter adjustment is considered. Preferably, in actual application, if the parameter smoothness penalty term is not considered, the control parameters will be frequently and greatly adjusted, which will cause the pantograph-catenary system to vibrate intensively and accelerate the wear. By balancing the reward and the penalty, the system can effectively reduce the wear offset under the premise of stable operation, and the practicability and effectiveness of the pantograph-catenary wear intelligent detection and control scheme are significantly improved.

[0036] Further, the first fatigue crack propagation path set under a plurality of control parameter combinations and the second fatigue crack propagation path set under a plurality of control parameter combinations are obtained by respectively performing finite element simulation inversion, and the method further comprises the following steps:

[0037] Based on the pantograph-raising mechanism control parameter and the pantograph-lowering mechanism control parameter, a first control parameter combination matrix and a second control parameter combination matrix are respectively set; and the first control parameter combination matrix and the second control parameter combination matrix are configured based on an orthogonal table to obtain the plurality of control parameter combinations.

[0038] Specifically, based on the control parameters of the pantograph-raising mechanism and the pantograph-lowering mechanism, a first control parameter combination matrix and a second control parameter combination matrix are respectively set, which means that two different parameter combination matrices are constructed according to different control parameters in the pantograph-raising and pantograph-lowering processes, for subsequent simulation and optimization analysis; and the first control parameter combination matrix and the second control parameter combination matrix are configured based on an orthogonal table, so as to obtain a plurality of representative control parameter combinations by using the orthogonal experimental design method, for evaluating the influence of different parameter combinations on the pantograph fatigue crack propagation path.

[0039] The first control parameter combination matrix and the second control parameter combination matrix are set, wherein the first control parameter combination matrix corresponds to the control parameters of the pantograph-raising mechanism, such as air pressure value and pantograph-raising speed, and a parameter combination matrix is constructed, and these parameters have important influence on the stress distribution and crack propagation of the pantograph in the pantograph-raising process; and the second control parameter combination matrix corresponds to the control parameters of the pantograph-lowering mechanism, such as lowering angle and damping coefficient, and a parameter combination matrix is constructed, and these parameters also have important influence on the stress distribution and crack propagation of the pantograph in the pantograph-lowering process.

[0040] In the control optimization model of the pantograph, the control parameter variables include gain coefficients and integral coefficients, which need to be optimized through the form of combination matrix; representative parameter combinations are selected from the combination matrix of class I and class II control parameters using an orthogonal table. The orthogonal table design can ensure that all possible combinations of parameters are covered in a limited number of experiments, improving the efficiency of the experiment and the reliability of the results. Specifically, when performing statics optimization of the pantograph, the combination of multiple parameters needs to be considered, and the key parameter combinations can be effectively screened out through orthogonal table configuration.

[0041] Through orthogonal table configuration, multiple different control parameter combinations are obtained, which will be used for subsequent finite element simulation inversion to analyze the fatigue crack propagation path of the pantograph under different working conditions. In the above steps, by setting two types of control parameter combination matrix and performing orthogonal table configuration, the influence of different parameter combinations on the fatigue crack propagation of the pantograph can be systematically explored, providing a scientific basis for subsequent simulation analysis and optimization, ensuring the comprehensiveness and reliability of the results, and helping to more accurately assess the wear of the pantograph, so as to develop a reasonable maintenance strategy.

[0042] Further, the method of the present application further comprises:

[0043] Based on the pantograph mechanism control parameters, a first parameter optimization space is constructed; the partial derivatives of the crack propagation rate with respect to the air pressure value and the pantograph raising speed in the pantograph mechanism control parameters are calculated, and redundant control parameter combinations in the low sensitivity region are determined; in the first parameter optimization space, the redundant control parameter combinations are used for redundancy elimination, and a class I control parameter combination matrix is configured.

[0044] Specifically, constructing a first parameter optimization space based on the pantograph mechanism control parameters means establishing a parameter optimization range and framework according to the control parameters of the pantograph mechanism (such as air pressure value, pantograph raising speed, etc.), which is used for subsequent optimization analysis. Calculating the partial derivatives of the crack propagation rate with respect to the air pressure value and the pantograph raising speed in the pantograph mechanism control parameters means determining the sensitivity of the crack propagation rate to these control parameters through mathematical methods. Determining redundant control parameter combinations in the low sensitivity region means finding parameter combinations that have less influence on the crack propagation rate. Using these redundant control parameter combinations for redundancy elimination in the first parameter optimization space and configuring a class I control parameter combination matrix means removing these redundant parameter combinations from the optimization space, thereby obtaining a more concise and effective control parameter combination matrix, which is used for subsequent analysis and optimization.

[0045] According to the control parameters (such as air pressure value, bow lifting speed, etc.) of the bow lifting mechanism, the reasonable range and value step of the parameters are determined, a parameter optimization space is constructed, which covers all possible combinations of control parameters, and provides a basis for subsequent optimization analysis; the partial derivatives of the crack propagation rate with respect to the air pressure value and the bow lifting speed are obtained, and the influence of these parameters on the crack propagation rate is obtained, and by analyzing the size of the partial derivative, it is determined which parameter combination is in the low sensitivity area, that is, the combination with less influence on the crack propagation rate. Specifically, in the control optimization model of the pantograph, by constructing a PD controller of the pantograph-catenary contact force, the influence of different control parameter combinations on the output error of the contact force can be calculated, so as to determine the redundant parameter combination.

[0046] In the first parameter optimization space constructed, the redundant control parameter combination is removed, and the parameter combination with greater influence on the crack propagation rate is retained to form a control parameter combination matrix, which improves the pertinence and effectiveness of the parameter combination and reduces unnecessary calculation. In the above steps, by constructing the parameter optimization space and calculating the partial derivative, the influence of the control parameters of the bow lifting mechanism on the crack propagation rate can be systematically analyzed, which provides a scientific basis for subsequent simulation and optimization. The control parameter combination matrix obtained after removing the redundant parameter combination not only improves the calculation efficiency, but also ensures the accuracy and reliability of the analysis results, which helps to more accurately evaluate the wear of the pantograph, so as to develop a reasonable maintenance strategy.

[0047] Further, the method of the application further comprises:

[0048] The crack propagation rate function is defined as: The crack propagation rate The first-order partial derivative of the air pressure value in the control parameters of the bow lifting mechanism The crack propagation rate The first-order partial derivative of the bow lifting speed in the control parameters of the bow lifting mechanism , wherein is used to represent the crack length, is used to represent the cycle number.

[0049] Specifically, the crack propagation rate function is defined as: , wherein is the crack propagation rate, is the crack propagation rate, the first-order partial derivative of the crack propagation rate with respect to the air pressure value in the control parameters of the bow lifting mechanism , the first-order partial derivative of the bow lifting speed in the control parameters of the bow lifting mechanism The sensitivity of the crack propagation rate to these control parameters, the variables of the crack length and the cycle number, are used to quantify the crack propagation and the fatigue life.

[0050] According to the material mechanics and the fatigue crack propagation theory, the crack propagation rate function is established, which is expressed as: wherein, is used to represent the crack length, is used to represent the cycle number; the first-order partial derivatives of the crack propagation rate function with respect to the air pressure value and the bowing speed are calculated, and the results are: , which represents the sensitivity of the crack propagation rate to the air pressure value and the bowing speed. By calculating these partial derivatives, it can be determined which control parameters have a greater impact on the crack propagation rate. In the above steps, by defining the crack propagation rate function and calculating the partial derivatives, the sensitivity of the crack propagation rate to the control parameters of the bowing mechanism is quantified, which provides a scientific basis for subsequent parameter optimization and redundancy elimination, and helps to more accurately assess the fatigue life and wear of the pantograph.

[0051] Further, the method of the present application further comprises:

[0052] The second-order partial derivatives of the crack propagation rate with respect to the air pressure value and the bowing speed of the control parameters of the bowing mechanism are calculated: Based on the second-order partial derivatives of the air pressure value and the bowing speed , the interaction effect under the second-order partial derivatives is captured, and the boundary of the low sensitivity region is adaptively adjusted.

[0053] Specifically, the second-order partial derivatives of the crack propagation rate with respect to the air pressure value and the bowing speed of the control parameters of the bowing mechanism are calculated based on the crack propagation rate function. These second-order partial derivatives can capture the interaction effect of the crack propagation rate on the control parameters, i.e. the comprehensive influence of the two parameters on the crack propagation rate when they change simultaneously. Based on these second-order partial derivatives, the boundary of the low sensitivity region is adaptively adjusted, i.e. according to the results of the second-order partial derivatives, the range of parameter combinations that have a smaller impact on the crack propagation rate is dynamically adjusted, so as to optimize the parameter combination matrix.

[0054] Based on the crack propagation rate function , the second-order partial derivatives of the crack propagation rate with respect to the air pressure value and the bowing speed are calculated: These second-order partial derivatives can capture the interaction effect of crack propagation rate on air pressure value and arching speed, that is, the comprehensive influence of the two parameters on the crack propagation rate when they change simultaneously; by analyzing the size and sign of the second-order partial derivative, the interaction effect between the air pressure value and the arching speed is determined, if is positive, it means that the increase of air pressure value and arching speed will jointly promote the increase of crack propagation rate; if is negative, it means that the increase of air pressure value and arching speed will jointly inhibit the increase of crack propagation rate.

[0055] According to the results of the second-order partial derivative, the boundary of the low sensitivity region is adaptively adjusted, specifically, if a parameter combination is in the low sensitivity region and its second-order partial derivative indicates that the region has less influence on the crack propagation rate, the boundary of the region can be appropriately expanded to reduce unnecessary calculation and analysis. Specifically, by dynamically adjusting the range of parameter combinations, the analysis is focused on the regions that have greater influence on the crack propagation rate, thereby improving the calculation efficiency and the accuracy of the results.

[0056] By calculating the second-order partial derivative and capturing the interaction effect, the sensitivity of the crack propagation rate to the control parameters of the arching mechanism can be more comprehensively understood, providing a more detailed basis for subsequent parameter optimization and redundancy elimination. The adaptive adjustment of the boundary further optimizes the parameter combination matrix, ensuring the efficiency and relevance of the analysis, which helps to more accurately assess the wear of the pantograph, thereby developing a reasonable maintenance strategy and improving the reliability and safety of the system.

[0057] Further, the method of the application further comprises:

[0058] Based on the control parameters of the arching mechanism, a second parameter optimization space is constructed; a damping coefficient fluctuation range and a lowering angle change interval are introduced; in the first parameter optimization space, data filtering is performed using the damping coefficient fluctuation range and the lowering angle change interval, and a second type of control parameter combination matrix is configured.

[0059] Specifically, constructing a second parameter optimization space based on the control parameters of the arching mechanism means that on the basis of the previously constructed first parameter optimization space, other related control parameters such as damping coefficient and lowering angle are further considered to construct a more comprehensive parameter optimization range. The introduction of the damping coefficient fluctuation range and the lowering angle change interval is to consider the changes of these parameters in actual work, so that the optimization result is more in line with the actual situation. In the first parameter optimization space, data filtering is performed using these intervals, and a second type of control parameter combination matrix is configured, which means that parameter combinations that meet the requirements of the damping coefficient fluctuation range and the lowering angle change interval are selected from the first parameter optimization space to form a new control parameter combination matrix for subsequent analysis and optimization.

[0060] On the basis of the existing pantograph mechanism control parameters, further consider the damping coefficient and the lowering angle parameters, build a more comprehensive second parameter optimization space, this space covers all possible control parameter combinations, including air pressure value, pantograph speed, damping coefficient and lowering angle, etc., for the subsequent optimization analysis provides a more comprehensive basis, specifically, in the integrated optimization model of high-speed pantograph, need to consider multiple factors, including rod length parameters, section parameters and control parameters, etc., to build a comprehensive optimization space, to ensure the performance of the pantograph is fully improved.

[0061] The damping coefficient fluctuation range refers to the possible change interval of the damping coefficient during the operation of the pantograph, and the lowering angle change interval refers to the possible change range of the lowering angle during the lowering process. The introduction of these parameters makes the optimization model more close to the actual working condition and better reflects the performance of the pantograph under different working conditions. Specifically, in the dynamic optimization model of the pantograph, the influence of damping coefficient and lowering angle on the dynamic performance of the pantograph needs to be considered. By introducing the change interval of these parameters, the performance of the pantograph can be more comprehensively evaluated.

[0062] In the first parameter optimization space, according to the introduced damping coefficient fluctuation range and lowering angle change interval, the original parameter combinations are screened and filtered, and those that meet the requirements of these intervals are retained, and the combinations that do not meet the requirements are eliminated, forming a two-class control parameter combination matrix, which improves the pertinence and effectiveness of the parameter combinations, reduces unnecessary calculation and analysis, and makes the subsequent optimization process more efficient and accurate. For example, in the optimization algorithm, data filtering can ensure that the analysis focuses on parameter combinations that have a greater impact on the performance of the pantograph, thereby improving the reliability and practicality of the optimization results.

[0063] By constructing the second parameter optimization space and introducing the damping coefficient fluctuation range and lowering angle change interval, the performance of the pantograph under different working conditions can be more comprehensively considered. The two-class control parameter combination matrix after data filtering not only improves the calculation efficiency, but also ensures the accuracy and reliability of the analysis results, which helps to more accurately evaluate the wear of the pantograph, so as to develop a reasonable maintenance strategy and improve the reliability and safety of the system.

[0064] Further, the first fatigue crack propagation path set and the second fatigue crack propagation path set are subjected to three-dimensional convolution processing to extract spatiotemporal correlation features. The method of the present application includes:

[0065] The gradient features of the extended paths are extracted along the time, space, and crack depth dimensions, and a three-dimensional convolution kernel is set; based on the three-dimensional convolution kernel, the first fatigue crack extension path set and the second fatigue crack extension path set are subjected to frequency domain decomposition, and multi-band features are extracted, the multi-band features including high-frequency mutation features and low-frequency trend features; based on the multi-band features, a spatio-temporal correlation feature is determined by using an attention gate mechanism.

[0066] Specifically, the gradient features of the extended paths are extracted along the time, space, and crack depth dimensions, which means that the gradient is calculated on the crack extension path in the time dimension (cycle number), the space dimension (crack position), and the crack depth dimension, respectively, to capture the change trend and direction of crack extension, and a three-dimensional convolution kernel is set, which means that a three-dimensional convolution kernel is constructed for convolution operation in three-dimensional space to extract multi-dimensional features, and the frequency domain decomposition is performed based on the three-dimensional convolution kernel, which means that the crack extension path set is converted from the time domain to the frequency domain to analyze the features under different frequency components, and the multi-band features include high-frequency mutation features and low-frequency trend features, which respectively reflect the rapid change and slow change trend of crack extension, and the spatio-temporal correlation feature is determined by using an attention gate mechanism, which means that the multi-band features extracted are weighted by using the attention mechanism to highlight important spatio-temporal features and suppress unimportant features, thereby improving the attention of the model to key information.

[0067] On the crack extension path, the gradient features are extracted along the time, space, and crack depth dimensions, for example, in the analysis of the spatio-temporal features of the traffic flow network, the features are extracted along multiple dimensions such as time and space to capture the dynamic changes of the traffic flow, and a three-dimensional convolution kernel is set, which has a size of 3×3×3 or 5×5×5, for convolution operation in three-dimensional space to extract multi-dimensional features, and the three-dimensional convolution kernel can capture the feature changes in the time, space, and depth dimensions simultaneously, thereby improving the comprehensiveness and accuracy of feature extraction.

[0068] Based on the three-dimensional convolution kernel, the first fatigue crack extension path set and the second fatigue crack extension path set are subjected to frequency domain decomposition, for example, in the processing of the acoustic emission signals of fatigue crack extension, the frequency domain decomposition is performed by using wavelet transform or other methods to extract the features under different frequency components; the multi-band features are extracted, including high-frequency mutation features and low-frequency trend features, the high-frequency mutation features reflect the rapid changes and local mutations in the crack extension process, and the low-frequency trend features reflect the overall trend and slow changes of the crack extension.

[0069] Based on the extracted multi-band features, an attention gate mechanism is constructed, for example, in a prediction model with an integrated attention mechanism, an attention mechanism is constructed through a global average pooling layer, a fully connected layer and a sigmoid layer, etc. to weight the features. Through the attention gate mechanism, the spatio-temporal correlation features are determined. The attention mechanism can highlight important spatio-temporal features and suppress unimportant features, so that the model pays more attention to the key areas and time periods of crack propagation, and improves the accuracy and robustness of the model.

[0070] By extracting gradient features of the propagation path along the time, space and crack depth dimensions, and setting a three-dimensional convolution kernel, the multi-dimensional features of crack propagation can be comprehensively captured. Frequency domain decomposition and multi-band feature extraction further analyze the different frequency components of crack propagation, providing a basis for fine description of features. The attention gate mechanism is used to determine the spatio-temporal correlation features, which improves the attention of the model to key information, enhances the feature expression ability and prediction accuracy of the model, and helps to more accurately evaluate the wear condition of the pantograph, so as to develop a reasonable maintenance strategy.

[0071] In summary, the beneficial effects of the embodiments of the present application are:

[0072] Since the pantograph working state and surface image information are obtained in the pantograph-catenary contact area between the catenary wire and the slide plate, the control parameters of the pantograph-raising mechanism corresponding to the upper frame and the lower arm, and the control parameters of the pantograph-lowering mechanism corresponding to the upper frame and the lower arm are respectively subjected to finite element simulation inversion to obtain a first set of fatigue crack propagation paths and a second set of fatigue crack propagation paths under a plurality of control parameter combinations, and the spatio-temporal correlation features are extracted through three-dimensional convolution processing. Based on the pantograph working state and surface image information, environmental compensation data is added to correct the perception deviation, and multi-scale feature fusion is performed in combination with the first set of fatigue crack propagation paths and the second set of fatigue crack propagation paths to generate a dynamic wear feature tensor under the spatio-temporal correlation features for wear offset control. The present application provides a pantograph-catenary wear intelligent detection method and system based on machine vision, which inverses the fatigue crack propagation paths of the pantograph-raising mechanism and the pantograph-lowering mechanism, comprehensively evaluates the wear condition of the pantograph-catenary, remotely detects and maintains the pantograph and the catenary in real time, makes maintenance decisions, discovers and handles potential wear hazards in a timely manner, avoids accidents, and improves the technical effects of detection accuracy and efficiency. Embodiments

[0073] Based on the same inventive concept as the pantograph-catenary wear intelligent detection method based on machine vision in the foregoing embodiments, as Figure 2 shown, the embodiments of the present application provide a pantograph-catenary wear intelligent detection system based on machine vision, wherein the system comprises:

[0074] An information acquisition module M100 is configured to acquire pantograph working state and surface image information in an overhead line-pantograph contact area between the overhead line and the pantograph.

[0075] An emulation module M200 is configured to respectively perform finite element emulation inversion on pantograph-raising mechanism control parameters corresponding to the upper frame and the lower arm, and pantograph-lowering mechanism control parameters corresponding to the upper frame and the lower arm, to obtain a first fatigue crack propagation path set under a plurality of control parameter combinations and a second fatigue crack propagation path set under the plurality of control parameter combinations.

[0076] A convolution processing module M300 is configured to perform three-dimensional convolution processing on the first fatigue crack propagation path set and the second fatigue crack propagation path set to extract spatiotemporal correlation features.

[0077] A feature fusion module M400 is configured to add environmental compensation data to correct perception deviation based on the pantograph working state and the surface image information, and perform multi-scale feature fusion on the first fatigue crack propagation path set and the second fatigue crack propagation path set to generate a dynamic wear feature tensor under spatiotemporal correlation features.

[0078] An offset control module M500 is configured to perform wear offset control based on the dynamic wear feature tensor and in combination with the pantograph-raising mechanism control parameters and the pantograph-lowering mechanism control parameters.

[0079] Further, the offset control module M500 is configured to perform the following method:

[0080] Based on the dynamic wear feature tensor, determine a wear abnormal area and a wear depth evolution index; and update a feasible region of the pantograph-raising mechanism control parameters and the pantograph-lowering mechanism control parameters according to the wear abnormal area and the wear depth evolution index, the feasible region being mapped to a parameter combination with minimized wear offset.

[0081] Further, the offset control module M500 is further configured to perform the following method:

[0082] Perform iterative search in the updated feasible region, with minimized wear offset as a reward function; and in the iterative search process, smoothness of the pantograph-raising mechanism control parameters and smoothness of the pantograph-lowering mechanism control parameters as penalty terms.

[0083] Further, the emulation module M200 is further configured to perform the following method:

[0084] Based on the pantograph-raising mechanism control parameters and the pantograph-lowering mechanism control parameters, set a first control parameter combination matrix and a second control parameter combination matrix; and based on the first control parameter combination matrix and the second control parameter combination matrix, perform orthogonal table configuration to obtain the plurality of control parameter combinations.

[0085] Further, the simulation module M200 is further configured to perform the following method:

[0086] Based on the arch raising mechanism control parameters, a first parameter optimization space is constructed; partial derivatives of the crack propagation rate with respect to the air pressure value and the arch raising speed in the arch raising mechanism control parameters are calculated, and a redundant control parameter combination in a low sensitivity region is determined; in the first parameter optimization space, the redundant control parameter combination is used for redundancy elimination, and a first control parameter combination matrix is configured.

[0087] Further, the simulation module M200 is further configured to perform the following method:

[0088] A crack propagation rate function is defined: The crack propagation rate The first-order partial derivative of the air pressure value in the arch raising mechanism control parameters The crack propagation rate The first-order partial derivative of the arch raising speed in the arch raising mechanism control parameters Wherein, is used to represent the crack length, is used to represent the cycle number.

[0089] Further, the simulation module M200 is further configured to perform the following method:

[0090] The second-order partial derivative of the crack propagation rate with respect to the air pressure value and the arch raising speed in the arch raising mechanism control parameters: Based on the second-order partial derivatives of the air pressure value and the arch raising speed , the interaction effect under the second-order partial derivative is captured, and the low sensitivity region is adaptively adjusted.

[0091] Further, the simulation module M200 is further configured to perform the following method:

[0092] Based on the arch raising mechanism control parameters, a second parameter optimization space is constructed; a damping coefficient fluctuation range and a lowering angle change interval are introduced; in the first parameter optimization space, the damping coefficient fluctuation range and the lowering angle change interval are used for data filtering, and a second control parameter combination matrix is configured.

[0093] Further, the convolution processing module M300 is configured to perform the following method:

[0094] The gradient features of the extended paths are extracted along the time, space and crack depth dimensions, and a three-dimensional convolution kernel is set; based on the three-dimensional convolution kernel, the first fatigue crack extension path set and the second fatigue crack extension path set are subjected to frequency domain decomposition, and multi-band features are extracted, the multi-band features including high-frequency mutation features and low-frequency trend features; based on the multi-band features, a spatio-temporal correlation feature is determined by using an attention gate mechanism.

[0095] In summary, any step can be stored as computer instructions or programs in an unrestricted computer memory and can be called and recognized by an unrestricted computer processor, and no unnecessary limitation is made here.

[0096] Further, the above technical solutions only represent preferred technical solutions of the technical solutions of the embodiments of the present application, and some variations of certain parts made by those skilled in the art also represent the principles of the novel embodiments of the present application. Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the scope of the present application.

Claims

1. A machine vision-based intelligent detection method for pantograph-catenary wear, characterized in that, The method includes: In the pantograph-catenary contact area between the contact wire and the sliding plate, obtain pantograph working status and surface image information; The control parameters of the lifting mechanism corresponding to the upper frame and the lower boom, and the control parameters of the lowering mechanism corresponding to the upper frame and the lower boom, were respectively subjected to finite element simulation inversion to obtain the first fatigue crack propagation path set and the second fatigue crack propagation path set under multiple sets of control parameter combinations. The first fatigue crack propagation path set and the second fatigue crack propagation path set are subjected to three-dimensional convolution processing to extract spatiotemporal correlation features; Based on the pantograph's working status and surface image information, environmental compensation data is added to correct perception bias, and multi-scale feature fusion is performed by combining the first fatigue crack propagation path set and the second fatigue crack propagation path set to generate a dynamic wear feature tensor under spatiotemporal correlation features. Wear offset control is performed based on the dynamic wear characteristic tensor and the control parameters of the bow raising mechanism and bow lowering mechanism.

2. The intelligent detection method for bow and fire protection cable wear based on machine vision as described in claim 1, characterized in that, Based on the dynamic wear characteristic tensor, wear offset control is performed in conjunction with the control parameters of the raising and lowering mechanisms, including: Based on the dynamic wear feature tensor, the wear anomaly region and wear depth evolution index are determined; Based on the abnormal wear region and wear depth evolution index, the feasible region of the control parameters of the raising mechanism and the lowering mechanism is updated. The feasible region is mapped to the combination of control parameters that minimizes the wear offset obtained from the search.

3. The intelligent detection method for bow and catenary wear based on machine vision as described in claim 2, characterized in that, Perform an iterative search within the updated feasible region, using the minimization of wear offset as the reward function; During the iterative search process, the smoothness of the control parameters of the raising mechanism and the smoothness of the control parameters of the lowering mechanism are used as penalty terms.

4. The intelligent detection method for bow and catenary wear based on machine vision as described in claim 1, characterized in that, Finite element simulation inversions were performed to obtain the first fatigue crack propagation path set under multiple combinations of control parameters, the second fatigue crack propagation path set under multiple combinations of control parameters, and also including: Based on the control parameters of the raising and lowering mechanisms, a first-class control parameter combination matrix and a second-class control parameter combination matrix are respectively set. Based on the first type of control parameter combination matrix and the second type of control parameter combination matrix, an orthogonal table is configured to obtain the multiple sets of control parameter combinations.

5. The intelligent detection method for bow and catenary wear based on machine vision as described in claim 4, characterized in that, Based on the control parameters of the bow lifting mechanism, a first parameter optimization space is constructed; Calculate the partial derivatives of the crack propagation rate with respect to the air pressure value and the lifting speed in the control parameters of the lifting mechanism, and determine the combination of redundant control parameters in the low-sensitivity region; In the first parameter optimization space, the redundant control parameter combination is used to eliminate redundancy and configure a control parameter combination matrix.

6. The intelligent detection method for bow and catenary wear based on machine vision as described in claim 5, characterized in that, Define the crack propagation rate function: ; The crack propagation rate The air pressure value in the control parameters of the bow lifting mechanism First-order partial derivatives ; The crack propagation rate The bow lifting speed in the control parameters of the bow lifting mechanism First-order partial derivatives ,in, Used to characterize crack length Used to characterize the number of cycles.

7. The intelligent detection method for pantograph-catenary wear based on machine vision as described in claim 6, characterized in that, The crack propagation rate The air pressure value in the control parameters of the bow lifting mechanism and draw speed Second-order partial derivatives: , , ; Based on the air pressure value and draw speed The second-order partial derivatives are used to capture the interaction effects under the second-order partial derivatives, and the boundary adaptive adjustment is performed on the low-sensitivity region.

8. The intelligent detection method for bow and catenary wear based on machine vision as described in claim 4, characterized in that, Based on the control parameters of the bow lowering mechanism, a second parameter optimization space is constructed; Introduce the damping coefficient fluctuation range and the bow angle variation range; In the first parameter optimization space, the damping coefficient fluctuation range and the bow angle change range are used for data filtering, and a combination matrix of two types of control parameters is configured.

9. The intelligent detection method for bow and catenary wear based on machine vision as described in claim 1, characterized in that, The first fatigue crack propagation path set and the second fatigue crack propagation path set are subjected to three-dimensional convolution processing to extract spatiotemporal correlation features, including: Gradient features of the propagation path are extracted along the time, space, and crack depth dimensions, and a three-dimensional convolution kernel is set. Based on the three-dimensional convolution kernel, the first fatigue crack propagation path set and the second fatigue crack propagation path set are decomposed in the frequency domain to extract multi-frequency band features, including high-frequency abrupt change features and low-frequency trend features. Based on the multi-band features, spatiotemporal correlation features are determined using an attention gating mechanism.

10. A machine vision-based intelligent detection system for pantograph-catenary wear, characterized in that, For implementing the machine vision-based intelligent detection method for bow and cable rebar wear according to any one of claims 1-9, the system comprises: The information acquisition module is used to acquire pantograph working status and surface image information in the pantograph-catenary contact area between the contact wire and the sliding plate; The simulation module is used to perform finite element simulation inversion on the control parameters of the lifting mechanism corresponding to the upper frame and the lower boom, and the control parameters of the lowering mechanism corresponding to the upper frame and the lower boom, respectively, to obtain the first fatigue crack propagation path set and the second fatigue crack propagation path set under multiple sets of control parameter combinations. The convolution processing module is used to perform three-dimensional convolution processing on the first fatigue crack propagation path set and the second fatigue crack propagation path set to extract spatiotemporal correlation features. The feature fusion module is used to add environmental compensation data to correct perception deviations based on the pantograph's working state and surface image information, and to perform multi-scale feature fusion by combining the first fatigue crack propagation path set and the second fatigue crack propagation path set to generate a dynamic wear feature tensor under spatiotemporal correlation features. The offset control module is used to perform wear offset control based on the dynamic wear feature tensor and in combination with the control parameters of the bow lifting mechanism and the bow lowering mechanism.

Images