Method and system for monitoring a trolley brush, electronic device and storage medium

Abstract

A method and system for monitoring a contact wire collector brush, an electronic device and a storage medium, relate to the technical field of railway brush monitoring. The method comprises: acquiring a visible light image and a thermal infrared image of the brush at the current time; identifying the geometric contour of the brush according to the visible light image, calculating the current residual thickness according to the geometric contour, and calculating the initial abnormal value according to the current residual thickness; determining the current exchange area according to the visible light image, screening out a target sub-area located in the current exchange area, and extracting a target temperature value of the target sub-area; determining an abnormal area, adding the areas of each abnormal area to obtain an abnormal cumulative area, adjusting the initial abnormal value according to the abnormal cumulative area to obtain a target abnormal value; and determining that the brush is abnormal when the target abnormal value is higher than a preset abnormal threshold. The technical solution provided by the present application can improve the accuracy of the abnormal judgment of the brush.

Description

Technical Field

[0001] This application relates to the technical field of railway brush monitoring, specifically to a method, system, electronic device, and storage medium for monitoring the brushes of a contact current collector. Background Technology

[0002] In electrified railways or urban rail transit systems, the contact current collector is a key component for locomotives to obtain electrical energy from the overhead contact line. Its brushes slide directly in contact with the high-voltage contact wire, and while conducting current, they themselves wear down due to continuous friction and arc erosion. Since brushes are consumable parts, their wear condition directly affects the train's current-collecting stability and operational safety. Excessive wear or abnormal damage can lead to pantograph-catenary malfunctions and train interruptions, or even major safety accidents such as pantograph and contact line failure. Therefore, effective monitoring of brush wear is crucial.

[0003] Currently, the industry has proposed an online monitoring technology based on machine vision. This technology typically involves installing high-speed industrial cameras beside the track or on designated inspection gantry. When a train passes through the inspection area at normal speed, the camera automatically captures high-resolution images of the pantograph brush area. Subsequently, the backend image processing system analyzes the acquired images, using algorithms such as edge detection and feature recognition to accurately identify the brush outline in the image and compare it with a preset initial brush model to calculate the current remaining thickness of the brush. This non-contact online inspection method enables automated and dynamic measurement of brush geometry, improving inspection efficiency and allowing for a certain degree of prediction of wear trends.

[0004] However, while the aforementioned machine vision-based online monitoring technology can accurately acquire the macroscopic geometric wear of the brushes, it does not rely solely on thickness to assess the overall health of the brushes. During operation, brushes are subjected to complex coupling effects of multiple physical fields, including electrical, thermal, and mechanical forces. Especially during high-speed train operation or when the pantograph-catenary contact is poor, instantaneous arcing can cause a rapid increase in temperature in localized areas of the brush. While this high-temperature impact may not immediately result in a significant thickness change detectable by the vision system, it can cause qualitative changes in the microstructure of the carbon brush material, such as thermal stress cracking, material embrittlement, or deterioration of conductivity. This can lead to a dangerous situation where a brush's remaining thickness is within acceptable limits, but its material properties are severely degraded, making it susceptible to breakage or chipping under the impact and vibration of high-speed operation. Therefore, relying solely on thickness measurements can result in inaccurate assessments of brush anomalies. Summary of the Invention

[0005] This application provides a method, system, electronic device, and storage medium for monitoring the brushes of a contact current collector, which can improve the accuracy of brush anomaly detection.

[0006] The first aspect of this application provides a method for monitoring the brush of a contact current collector, specifically including:

[0007] When the train is in operation, and the brush is in contact with the contact wire and carrying the working current, a visible light image and a thermal infrared image of the brush are acquired at the same location at the current moment.

[0008] Based on the visible light image, the geometric contour of the brush is identified, and the current remaining thickness of the brush in the direction perpendicular to the contact wire is calculated based on the geometric contour. An initial outlier value is then calculated based on the current remaining thickness.

[0009] Based on the visible light image, the current exchange zone is determined. The current exchange zone is the continuous area where the brush and the contact wire are in contact at the current moment. Based on the thermal infrared image, the surface of the brush is divided into multiple sub-regions. The target sub-region located in the current exchange zone is selected, and the target temperature value of the target sub-region is extracted.

[0010] The target sub-regions whose target temperature values ​​are higher than the average temperature value of the current exchange region are identified as abnormal regions. The areas of the abnormal regions are added together to obtain the cumulative abnormal area. The initial abnormal value is adjusted according to the cumulative abnormal area to obtain the target abnormal value.

[0011] When the target abnormal value is higher than the preset abnormal threshold, it is determined that the brush has malfunctioned.

[0012] By employing the above technical solution, while acquiring visible light images of the brush to calculate the remaining thickness, thermal infrared images are simultaneously acquired to analyze the temperature distribution characteristics of the brush surface. This not only allows for the assessment of the brush's macroscopic wear state from a geometric perspective but also reflects changes in the brush material's microscopic properties through thermal imaging. Specifically, by identifying the current exchange zone and screening target sub-regions with abnormal temperatures within that zone, this invention effectively eliminates interference temperatures from non-working areas and accurately detects localized overheating phenomena caused by arc discharge, etc. By comprehensively considering the cumulative area of ​​abnormal regions with geometric wear, the obtained target anomaly value more comprehensively reflects the brush's health status, avoiding the risk of missed detections that might arise from relying solely on a single thickness indicator. This monitoring method based on multi-dimensional data fusion can promptly detect potential faults caused by material performance degradation, improving the accuracy of brush anomaly assessment.

[0013] Optionally, calculating the initial outlier based on the current remaining thickness includes:

[0014] Obtain the first remaining thickness sequence of the brush within a first preset time period before the current time, divide the first remaining thickness sequence into multiple time windows, and calculate the thickness change rate of the brush within each time window;

[0015] Based on the thickness change rate and the current remaining thickness, determine the first thickness change value of the brush within a second preset time period after the current moment;

[0016] Obtain the intermediate thickness sequence of the brush within a third preset time period before the current time. Based on the intermediate thickness sequence and the first thickness change value, predict the second thickness change value of the brush within a fourth preset time period after the current time. The third preset time period is greater than and includes the first preset time period, and the fourth preset time period is greater than and includes the second preset time period.

[0017] Calculate the weighted average of the first thickness change value and the second thickness change value, and divide the difference between the current remaining thickness and the weighted average by the weighted average to obtain the initial outlier value.

[0018] By employing the aforementioned technical solution, and through segmented processing and rate-of-change calculation of historical thickness data across different time spans, both short-term fluctuations and long-term evolution patterns of brush wear can be simultaneously captured. Specifically, the first thickness change value calculated based on the first remaining thickness sequence over a shorter time period reflects the recent wear state of the brush, while the second thickness change value predicted using the intermediate thickness sequence over a longer time period reflects the wear trend over a longer period. By weighted fusion of the prediction results from these two different time scales, both the real-time nature and stability of the prediction are ensured. The deviation between the current measured thickness and the predicted value is quantified as an initial outlier, providing reliable basic data for subsequent comprehensive evaluation combined with temperature characteristics, effectively improving the accuracy of brush anomaly identification.

[0019] Optionally, predicting the second thickness change value of the brush within a fourth preset time period after the current moment based on the intermediate thickness sequence and the first thickness change value includes:

[0020] Calculate the difference between two adjacent thickness values ​​in the intermediate thickness sequence, and form a difference sequence from each difference. Calculate the wear rate of the brush at each moment based on the difference sequence.

[0021] The moment when the wear rate is greater than the first rate threshold is defined as the abnormal wear moment, and the cumulative wear value within a preset wear time range before and after each abnormal wear moment is calculated;

[0022] The difference between adjacent cumulative wear values ​​is calculated, and the average of the differences between the cumulative wear values ​​is subtracted from the first thickness change value to obtain the second thickness change value of the brush within the fourth preset time period after the current moment.

[0023] By employing the aforementioned technical solution, and converting the difference sequence of adjacent thickness values ​​into wear rates, the abnormal moments of severe brush wear can be accurately identified. Furthermore, by calculating the cumulative wear value and its temporal evolution within a certain time range before and after the abnormal wear moment, not only can the impact of a single abnormal event be quantified, but the correlation between consecutive abnormal wear events can also be reflected. The impact of this abnormal wear pattern on the prediction results is corrected using the difference in cumulative wear values, making the prediction of the second thickness change value more consistent with the nonlinear wear characteristics of the brush in the actual operating environment, thereby improving the accuracy and reliability of the prediction. This prediction correction method based on historical abnormal wear characteristics can effectively address various abnormal operating conditions that brushes may encounter during actual operation, providing a more accurate basis for brush condition assessment.

[0024] Optionally, adjusting the initial outlier value based on the cumulative outlier area to obtain the target outlier value includes:

[0025] The area weighting coefficient of each abnormal region is obtained based on the proportion of the area of ​​each abnormal region to the total area of ​​abnormalities.

[0026] The abnormal region with the highest target temperature value is identified as the dominant abnormal region. Based on the area weighting coefficient of each abnormal region, the temperature decay rate of each abnormal region relative to the dominant abnormal region is calculated.

[0027] The target azimuth sector where each of the abnormal regions is located is determined, and the brush is composed of multiple preset azimuth sectors;

[0028] Calculate the dispersion of the temperature decay rate of the abnormal region in each of the target azimuth sectors, and determine the target azimuth sector with the largest dispersion as the abnormal propagation region;

[0029] The initial anomaly value is adjusted based on the temperature decay rate in the anomaly propagation region to obtain the target anomaly value.

[0030] By adopting the above technical solution, this application can achieve a deeper level of fault diagnosis. In particular, by dividing the brush surface into multiple directional sectors and analyzing the dispersion of temperature decay rate within each sector, it is possible to identify latent degradation caused by material inhomogeneity or internal microcracks distributed along specific directions. This directional diagnostic capability is not available in traditional monolithic hotspot analysis. Furthermore, by adjusting the initial anomaly value based on the characteristics of this anomaly propagation direction, the final evaluation result not only reflects wear and overall heat but also incorporates considerations of fault development trends and spatial distribution patterns, greatly improving the accuracy and predictability of the judgment.

[0031] Optionally, adjusting the initial anomaly value based on the temperature decay rate in the anomaly propagation region to obtain the target anomaly value includes:

[0032] Calculate the temperature decay rate difference between adjacent abnormal regions in the abnormal propagation region, and determine the temperature decay inflection point based on each temperature decay rate difference. The abnormal regions before the temperature decay inflection point are determined as the core abnormal region set.

[0033] The ratio of the average temperature decay rate within the core anomaly region set to the target temperature value of the dominant anomaly region is calculated to obtain the anomaly intensity coefficient.

[0034] The target anomaly value is obtained by multiplying the initial anomaly value by the reciprocal of the anomaly intensity coefficient.

[0035] By employing the aforementioned technical solution, analyzing the temperature decay rate difference between adjacent abnormal regions within the abnormal propagation area and identifying the temperature decay inflection point, the core abnormal region set with the most significant temperature anomalies can be accurately defined. This region division method based on temperature gradient changes avoids the judgment bias that may arise from simply using a fixed temperature threshold. Furthermore, by calculating the ratio of the average temperature decay rate within the core abnormal region set to the temperature of the dominant abnormal region, an anomaly intensity coefficient is obtained. This coefficient not only reflects the concentration of the abnormal temperature field but also the severity of the temperature anomaly. Adjusting the initial anomaly value using the reciprocal of the anomaly intensity coefficient allows the final target anomaly value to more accurately characterize the overall anomaly degree of the brush: when temperature anomalies are concentrated and decay slowly, it indicates potential severe material performance degradation, and the anomaly assessment value is increased accordingly; conversely, if temperature anomalies are dispersed and decay rapidly, it may only be a temporary overheating phenomenon, and the anomaly assessment value is decreased accordingly. This dynamic adjustment mechanism considering the temperature field distribution characteristics improves the accuracy and reliability of brush anomaly state identification.

[0036] Optionally, after determining that the brush has malfunctioned, the method further includes:

[0037] Based on the visible light image and the thermal infrared image, temporal and spatial data are constructed, which are data on the anomalies of the brush surface in time and space.

[0038] Based on the spatiotemporal data, the evolution rate of the abnormal region and the fluctuation frequency of the target temperature value are calculated.

[0039] The product of the evolution rate and the fluctuation frequency is determined as the dynamic anomaly index;

[0040] Determine the dynamic anomaly index sequence within a preset anomaly time period after the brush malfunctions, and calculate the remaining service life of the brush based on the dynamic anomaly index sequence.

[0041] When the changing trend of the dynamic anomaly index meets the preset early warning conditions, the optimal brush replacement time window is calculated based on the remaining service life.

[0042] Within the optimal brushing time window, the idle time point of the train is selected as the target brushing time according to the train operation plan.

[0043] By employing the aforementioned technical solutions, spatiotemporal data is constructed to quantify the spatiotemporal evolution characteristics of abnormal brush surface states, providing a complete data foundation for subsequent analysis. By calculating the evolution rate of the abnormal region and the fluctuation frequency of the target temperature value, and combining them into a dynamic anomaly index, not only can the development speed of the abnormal state be reflected, but also the stability of the anomaly degree can be demonstrated. Specifically, this invention establishes a predictive model for the remaining service life of the brush by analyzing the dynamic anomaly index sequence within a preset abnormal time period, and makes early warning judgments based on the changing trend of the dynamic anomaly index, achieving early prediction of brush failure risk. Furthermore, by calculating the optimal brush replacement time window and combining it with the actual train operation plan to select the target brush replacement time, both the technical rationality of the maintenance timing and the actual constraints of operational scheduling are considered, thereby achieving proactive prevention and precise scheduling of brush maintenance. This intelligent management scheme, which combines condition monitoring, life prediction, and maintenance decision-making, can effectively avoid safety hazards caused by premature brush replacement or operation with defects, significantly improving the operational efficiency and safety reliability of electrified railway systems.

[0044] Optionally, calculating the remaining service life of the brush based on the dynamic anomaly index sequence includes:

[0045] Calculate the mean of each dynamic anomaly index in the dynamic anomaly index sequence;

[0046] The time corresponding to the dynamic anomaly index that is greater than the mean is determined as a high-risk time. The absolute value of the difference of the dynamic anomaly index between each high-risk time is calculated to obtain the fluctuation range of the dynamic anomaly index.

[0047] The total duration of high-risk periods with fluctuations exceeding a preset fluctuation threshold is calculated, and the ratio of this total duration to the sum of all high-risk periods is determined as the lifespan coefficient.

[0048] The remaining service life of the brush is obtained by multiplying the preset standard lifespan of the brush by the lifespan coefficient.

[0049] By employing the aforementioned technical solution, calculating the mean of the dynamic anomaly index sequence and identifying high-risk moments, key evolution nodes of brush abnormalities can be effectively captured. Furthermore, by analyzing the fluctuation amplitude of the dynamic anomaly index between high-risk moments, not only can the time periods of drastic changes in abnormal states be identified, but the accelerating trend of brush performance degradation can also be reflected. Specifically, this invention obtains a lifespan coefficient by statistically analyzing the proportion of high-risk time periods with large fluctuations, establishing a lifespan prediction model that comprehensively considers the severity and duration of abnormal states. This model dynamically corrects the preset standard lifespan, ensuring that the predicted remaining lifespan accurately reflects the performance degradation pattern of the brush in the actual operating environment: frequent large fluctuations indicate that the brush may be in a rapid degradation stage, thus shortening the predicted lifespan accordingly; conversely, if the abnormal state is relatively stable, a relatively long predicted lifespan is maintained. This adaptive lifespan prediction method based on dynamic characteristics provides a scientific basis for timely brush replacement and preventative maintenance.

[0050] A second aspect of this application provides a contact current collector brush monitoring system, specifically comprising:

[0051] The image acquisition module is used to acquire a visible light image and a thermal infrared image of the brush at the same location when the brush is in contact with the contact wire and carrying the working current while the train is running.

[0052] The initial outlier calculation module is used to identify the geometric contour of the brush based on the visible light image, calculate the current remaining thickness of the brush in the direction perpendicular to the contact wire based on the geometric contour, and calculate the initial outlier based on the current remaining thickness.

[0053] The target temperature value extraction module is used to determine the current exchange zone based on the visible light image. The current exchange zone is the continuous area where the brush and the contact wire are pressed and contacted at the current moment. Based on the thermal infrared image, the surface of the brush is divided into multiple sub-regions, the target sub-region located in the current exchange zone is selected, and the target temperature value of the target sub-region is extracted.

[0054] The target anomaly calculation module is used to identify target sub-regions whose target temperature values ​​are higher than the average temperature value of the current exchange zone as anomaly regions, and to add the areas of each of the anomaly regions to obtain the cumulative anomaly area. The initial anomaly value is adjusted according to the cumulative anomaly area to obtain the target anomaly value.

[0055] The anomaly determination module is used to determine that the brush has malfunctioned when the target anomaly value is higher than a preset anomaly threshold.

[0056] A third aspect of this application provides an electronic device including a processor, a memory, a user interface, and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any of the foregoing.

[0057] A fourth aspect of this application provides a computer-readable storage medium storing instructions that, when executed, perform the method described in any of the preceding descriptions. Attached Figure Description

[0058] Figure 1 This is an exemplary system architecture diagram of a contact wire current collector brush monitoring system provided in an embodiment of this application;

[0059] Figure 2 This is a schematic flowchart of a contact current collector brush monitoring method provided in an embodiment of this application;

[0060] Figure 3 This is a schematic diagram of a directional sector division provided in an embodiment of this application;

[0061] Figure 4 This is a schematic diagram of the structure of a contact current collector brush monitoring system provided in an embodiment of this application;

[0062] Figure 5 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application.

[0063] Explanation of reference numerals in the attached figures: 901, processor; 902, communication bus; 903, user interface; 904, network interface; 905, memory. Detailed Implementation

[0064] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0065] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0066] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0067] Figure 1 A schematic diagram of an exemplary system architecture for a contact current collector brush monitoring system is shown.

[0068] like Figure 1 As shown, the system architecture may include a contact current collector brush monitoring device 011, a network 012, and an electronic device 013. The network 012 provides a data transmission link between the contact current collector brush monitoring device 011 and the electronic device 013. The network 012 may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0069] The contact current collector brush monitoring device 011 can interact bidirectionally with the electronic device 013 via network 012. The contact current collector brush monitoring device 011 is primarily responsible for acquiring visible light and thermal infrared images of the brush and reporting this image data to the electronic device 013 in real time. The contact current collector brush monitoring device 011 is hardware and may include image acquisition devices such as high-speed industrial cameras and thermal imagers, as well as embedded processing units for image preprocessing and data transmission.

[0070] Electronic Equipment 013 provides a brush monitoring and analysis platform responsible for processing and analyzing reported image data. The platform calculates the remaining brush thickness based on visible light images, analyzes temperature distribution characteristics using thermal infrared images, and achieves a comprehensive assessment of brush abnormalities through multi-dimensional data fusion. Electronic Equipment 013 can also perform lifespan prediction and maintenance decisions, providing optimized recommendations for brush replacement. These analytical results can be used for train operation safety monitoring and preventative maintenance management.

[0071] It should be noted that electronic devices can be either hardware or software. When an electronic device is hardware, it can be implemented as a distributed computing platform consisting of multiple servers to support the parallel processing of large-scale image data. When an electronic device is software, it can be implemented as multiple functional modules, including image processing modules, anomaly assessment modules, and lifetime prediction modules, etc. No specific limitations are made here.

[0072] It should be understood that Figure 1 The number of contact current collector brush monitoring devices 011, network 012, and electronic devices 013 shown in the diagram is merely illustrative. Depending on actual deployment needs, monitoring devices can be deployed simultaneously in multiple locations, with data aggregation and analysis conducted through a unified network access platform. Specifically, during the system debugging phase, the above system architecture may exclude network 012, and local connection methods may be used for development and testing.

[0073] The following description uses an electronic device as an example to illustrate a contact current collector brush monitoring method provided in this application.

[0074] This application provides a method for monitoring the brush of a contact current collector, referring to... Figure 2 , Figure 2 This is a flowchart illustrating a contact current collector brush monitoring method provided in an embodiment of this application, including steps S101 to S105, as follows:

[0075] S101: When the train is in operation, and the brush is in contact with the contact wire and carrying the working current, acquire the visible light image and thermal infrared image of the brush at the same location at the current moment.

[0076] In this embodiment, the visible light image refers to a digital image acquired by a high-speed industrial camera that reflects the appearance and wear state of the brush, used to characterize the brush's geometric features and surface condition. The thermal infrared image refers to a thermodynamic image captured by a thermal imager that reflects the temperature distribution on the brush surface, used to characterize the temperature field changes of the brush during the current collection process. These two types of images are acquired synchronously by imaging devices deployed at the same monitoring location, ensuring correlation analysis of multimodal characteristics of the same brush state.

[0077] Specifically, when the train is in operation and the brushes are in contact with the overhead contact line, the brushes, as key current-collecting components, carry a large current, which causes mechanical wear and temperature changes. The overhead contact line refers to the overhead metal conductor that provides traction current to the electric locomotive. It has a fixed suspension height and tension, used to maintain good contact with the brushes of the pantograph to achieve continuous power supply. To simultaneously capture the physical deformation and thermal characteristics of the brushes, high-speed industrial cameras and thermal imagers need to be deployed at fixed monitoring locations. The high-speed industrial camera faces the worn surface of the brush to acquire visible light images reflecting the brush's geometry; the thermal imager is aligned with the high-speed industrial camera in the same field of view to acquire thermal infrared images characterizing the brush's temperature distribution.

[0078] S102: Based on the visible light image, identify the geometric contour of the brush, calculate the current remaining thickness of the brush in the direction perpendicular to the contact wire based on the geometric contour, and calculate the initial outlier value based on the current remaining thickness.

[0079] In this embodiment, the initial outlier is a dimensionless index characterizing the brush state by comparing the current remaining thickness with the predicted change thickness (weighted average). Physically, the initial outlier reflects the degree of deviation of the current remaining thickness from the predicted change thickness. A zero initial outlier indicates that the current remaining thickness is roughly equal to the predicted change thickness; a positive initial outlier indicates that the current remaining thickness is greater than the predicted change thickness; and a negative initial outlier indicates that the current remaining thickness is less than the predicted change thickness. The larger the absolute value of the initial outlier, the greater the deviation of the current remaining thickness from the predicted change thickness.

[0080] Specifically, the acquired visible light images are first preprocessed, including image enhancement, noise reduction, and edge detection, to extract the geometric contours of the brushes. By establishing a mapping relationship between the image coordinate system and the actual physical dimensions, the current remaining thickness of the brushes perpendicular to the contact wire is calculated. To accurately assess the degree of brush anomalies, the thickness change trend of the brushes needs to be analyzed. First, the first remaining thickness sequence within a first preset time period (e.g., 24 hours) prior to the current moment is obtained. The first preset time period is chosen based on the daily operating characteristics of the brushes; setting it to 24 hours is to fully cover the wear data of a complete train operating cycle, reflecting the changes in the wear state of the brushes during different operating periods (e.g., morning peak, evening peak, off-peak). The first remaining thickness sequence is divided into multiple time windows (e.g., one window per hour), and the thickness change rate within each window is calculated. Based on these short-term thickness change rates and the current remaining thickness, the first thickness change value within a second preset time period (e.g., the next 12 hours) is predicted. The second preset time period is chosen to be 12 hours because brush wear has a relatively stable change trend within half an operating cycle; this duration is sufficient for short-term prediction without reducing accuracy due to an excessively long prediction period. Simultaneously, the intermediate thickness sequence is obtained within a longer third preset time period (e.g., 72 hours). The 72-hour (3-day) preset time period is based on analysis of actual operational data, which shows that brush wear exhibits cyclical changes over a 3-day timescale. This duration effectively smooths out the impact of short-term fluctuations without reducing system response speed due to excessive historical data. The second thickness change value is predicted within a fourth preset time period (e.g., the next 36 hours) based on the first thickness change value. The 36-hour preset time period balances prediction accuracy and early warning requirements, allowing sufficient preparation time for maintenance plans while maintaining the reliability of prediction results. Expanding the observation and prediction time windows improves prediction stability. Finally, the weighted average of the two predicted change values ​​is calculated, and the difference between the current remaining thickness and this average is divided by the average to obtain the initial anomaly value. By analyzing brush wear characteristics across four different time windows (24 hours, 12 hours, 72 hours, and 36 hours), both rapid response to short-term anomalies and reliable prediction of medium- to long-term degradation trends are ensured, forming a comprehensive spatiotemporal monitoring system. The selection of these time parameters is based on extensive experimental data verification and practical operational experience, and can be appropriately adjusted according to the characteristics of different routes and operational needs.

[0081] Based on the above embodiments, as an optional embodiment, S102: the step of calculating the initial outlier value based on the current remaining thickness may specifically include the following steps:

[0082] S201: Obtain the first remaining thickness sequence of the brush within the first preset time period before the current time, divide the first remaining thickness sequence into multiple time windows, and calculate the thickness change rate of the brush within each time window.

[0083] Specifically, a first preset time period (e.g., 4 hours) is selected before the current moment as the observation period. The remaining thickness data of the brush is continuously collected within the first preset time period to form a first remaining thickness sequence. The first remaining thickness sequence is divided into multiple time windows. For each time window, the thickness change rate within that time window is obtained by calculating the difference between the thickness at the end of the time window and the starting thickness, and dividing it by the duration of the time window.

[0084] S202: Based on the thickness change rate and the current remaining thickness, determine the first thickness change value of the brush within the second preset time period after the current moment.

[0085] In the embodiments of this application, the first thickness change value refers to the expected reduction in brush thickness over a future period of time based on short-term wear trends, and is used to characterize the recent wear development trend of the brush.

[0086] Specifically, firstly, by calculating the differences and patterns in the thickness change rate within each time window, the thickness change acceleration, reflecting the trend of thickness change, is obtained. Then, the thickness change acceleration is multiplied by a second preset time period (e.g., 24 hours) following the current moment to calculate the thickness change caused by the thickness change acceleration within the second preset time period. Next, the current remaining thickness is multiplied by the thickness change rate of the most recent time window to obtain the baseline change. Finally, the baseline change and the thickness change are added to obtain the first thickness change value considering the effect of acceleration.

[0087] S203: Obtain the intermediate thickness sequence of the brush within the third preset time period before the current moment, and predict the second thickness change value of the brush within the fourth preset time period after the current moment based on the intermediate thickness sequence and the first thickness change value. The third preset time period is greater than and includes the first preset time period, and the fourth preset time period is greater than and includes the second preset time period.

[0088] Specifically, the process first obtains the intermediate thickness sequence of the brush within a third preset time period (e.g., 72 hours) prior to the current moment. A difference sequence is formed by calculating the difference between two adjacent thickness values ​​in the sequence. Based on the difference sequence and the corresponding time intervals, the wear rate of the brush at each moment is calculated. A threshold judgment is applied to the calculated wear rate, and moments with wear rates exceeding a first threshold are marked as abnormal wear moments. For each abnormal wear moment, the thickness changes within a preset wear time range before and after it are analyzed, and the cumulative wear value is calculated. By calculating the difference between adjacent cumulative wear values ​​over time, the evolution characteristics of the abnormal wear process can be reflected. Finally, the average of all cumulative wear value differences is subtracted from the obtained first thickness change value to obtain the second thickness change value of the brush within a fourth preset time period (e.g., 7 days). It should be noted that the third preset time period is longer than and includes the first preset time period. This nested time design ensures a more reliable data foundation for short-term analysis, capturing sudden abnormal wear while avoiding misjudgments caused by short-term fluctuations. The fourth preset time period is longer than and includes the second preset time period. This design ensures that medium- and long-term forecasts maintain sensitivity to recent trends while making stable and reliable predictions based on data over a longer time span. This multi-layered time period design ensures data continuity across all forecasting stages and achieves a smooth transition from short-term to medium- and long-term forecasts, thereby improving the accuracy and reliability of the forecast results.

[0089] Based on the above embodiments, as an optional embodiment, S203: the step of predicting the second thickness change value of the brush within a fourth preset time period after the current moment according to the intermediate thickness sequence and the first thickness change value may specifically include the following steps:

[0090] S301: Calculate the difference between two adjacent thickness values ​​in the intermediate thickness sequence, and form a difference sequence from each difference. Calculate the wear rate of the brush at each moment based on the difference sequence.

[0091] Specifically, firstly, two adjacent thickness measurements are selected sequentially from the intermediate thickness sequence. The thickness difference between these two moments is obtained by subtracting the thickness value of the previous moment from the thickness value of the next moment. All thickness differences calculated between adjacent moment pairs are arranged in chronological order to form a complete difference sequence. Each value in the difference sequence corresponds to the actual wear of the brush within a specific time period. Subsequently, each difference in the difference sequence is divided by the corresponding time interval to obtain the instantaneous wear rate at each moment.

[0092] S302: Determine the moment when the wear rate is greater than the first rate threshold as the abnormal wear moment, and calculate the cumulative wear value within the preset wear time range before and after each abnormal wear moment.

[0093] Specifically, the calculated wear rate at each moment is first compared with a pre-set first rate threshold. When the wear rate at a certain moment exceeds the first rate threshold, that moment is marked as an abnormal wear moment. For each identified abnormal wear moment, a preset wear time range (e.g., 2 hours before and after) is determined. Within the determined time range, the thickness changes at each sampling moment are accumulated to obtain the cumulative wear value corresponding to that abnormal wear moment.

[0094] S303: Calculate the difference between adjacent cumulative wear values ​​in time, subtract the average of the differences between each cumulative wear value from the first thickness change value, and obtain the second thickness change value of the brush in the fourth preset time period after the current moment.

[0095] Specifically, firstly, the cumulative wear values ​​corresponding to each abnormal wear moment are arranged in chronological order, and the difference between two adjacent cumulative wear values ​​is calculated. The differences of all cumulative wear values ​​are then arithmetically averaged to obtain the overall mean. Finally, the first thickness change value is subtracted from the overall mean to obtain the second thickness change value of the brush within the fourth preset time period after the current moment (e.g., the next 7 days).

[0096] S204: Calculate the weighted average of the first thickness change value and the second thickness change value, and divide the difference between the current remaining thickness and the weighted average by the weighted average to obtain the initial outlier value.

[0097] Specifically, firstly, based on the different prediction time scales, corresponding weighting coefficients are assigned to the first and second thickness change values. The first and second thickness change values ​​are then multiplied by their respective weighting coefficients and summed to obtain a weighted average. Subsequently, the difference between the current remaining thickness and the weighted average is calculated, and this difference is divided by the weighted average to obtain the initial outlier.

[0098] S103: Based on the visible light image, determine the current exchange zone, which is the continuous area where the brush and the contact wire are pressed and in contact at the current moment. Based on the thermal infrared image, divide the brush surface into multiple sub-regions, screen out the target sub-region located in the current exchange zone, and extract the target temperature value of the target sub-region.

[0099] In the embodiments of this application, the current exchange region refers to the continuous area where the brush and the contact wire make squeezing contact and conduct current. It is used to characterize the actual working area of ​​the brush participating in current conduction. Its range directly affects the conductivity and heating characteristics of the brush and can be identified by the indentation features in the visible light image.

[0100] Specifically, firstly, image enhancement and edge detection are performed on the acquired visible light images. By analyzing the indentation morphology and gloss changes on the brush surface, continuous areas of compression contact with the contact wire are identified, and these areas are designated as the current exchange zone. Subsequently, the thermal infrared images are meshed, uniformly dividing the brush surface into multiple equal-sized sub-regions. By mapping the thermal infrared and visible light images to the same coordinate system, target sub-regions entirely within the current exchange zone are selected. The average or highest temperature of each target sub-region is then extracted as the target temperature value.

[0101] S104: The target sub-regions whose target temperature values ​​are higher than the average temperature value of the current exchange zone are identified as abnormal regions. The areas of each abnormal region are added together to obtain the cumulative abnormal area. The initial abnormal value is adjusted according to the cumulative abnormal area to obtain the target abnormal value.

[0102] In the embodiments of this application, the abnormal region refers to a local area where the temperature is significantly higher than the average level of the current exchange region. It is used to characterize the non-uniformity of the temperature distribution on the brush surface, and its distribution characteristics reflect the spatial pattern of local overheating or abnormal heating of the brush.

[0103] Specifically, firstly, the average temperature value of all target sub-regions within the current exchange zone is calculated. Target sub-regions with target temperatures exceeding this average temperature value are marked as anomalous regions, and the area of ​​each anomalous region is calculated and summed to obtain the cumulative anomalous area. Based on the proportion of each anomalous region's area to the cumulative anomalous area, a corresponding area weighting coefficient is calculated. Among all anomalous regions, the region with the highest target temperature value is identified as the dominant anomalous region. Based on the temperature value of the dominant anomalous region and the area weighting coefficient, the temperature decay rate of other anomalous regions relative to the dominant anomalous region is calculated.

[0104] The brush surface is pre-divided into multiple azimuth sectors, and the target azimuth sector to which each abnormal region belongs is determined. Within each target azimuth sector, the dispersion of the temperature decay rate of the abnormal region is analyzed, and the target azimuth sector with the largest dispersion is determined as the abnormal propagation region. Finally, based on the distribution characteristics of the temperature decay rate within the abnormal propagation region, an adjustment coefficient is established to correct the initial abnormal value, resulting in a more accurate target abnormal value reflecting the abnormal state of the brush.

[0105] Based on the above embodiments, as an optional embodiment, S104: the step of adjusting the initial outlier value according to the cumulative outlier area to obtain the target outlier value may specifically include the following steps:

[0106] S401: Based on the proportion of the area of ​​each abnormal region to the total area of ​​abnormalities, the area weight coefficient of each abnormal region is obtained.

[0107] Specifically, first, the area values ​​of all abnormal regions are obtained, and these area values ​​are summed to obtain the cumulative abnormal area. Then, the ratio of the area of ​​each abnormal region to the cumulative abnormal area is calculated to obtain the normalized area proportion. These area proportions are used as the area weight coefficients for the corresponding abnormal regions.

[0108] S402: Determine the abnormal region with the highest target temperature value as the dominant abnormal region, and calculate the temperature decay rate of each abnormal region relative to the dominant abnormal region by combining the area weight coefficient of each abnormal region.

[0109] In the embodiments of this application, the temperature decay rate refers to the degree to which the temperature value of each abnormal region decreases relative to the temperature value of the dominant abnormal region.

[0110] Specifically, firstly, the target temperature values ​​of all abnormal regions are compared, and the abnormal region with the highest temperature is identified as the dominant abnormal region. Then, the difference between the target temperature values ​​of other abnormal regions and the temperature value of the dominant abnormal region is calculated, and this difference is divided by the temperature value of the dominant abnormal region to obtain the base temperature decay rate. Multiplying the base temperature decay rate by the corresponding area weighting coefficient yields the temperature decay rate of each abnormal region relative to the dominant abnormal region.

[0111] S403: Determine the target azimuth sector where each abnormal area is located. The brush consists of multiple preset azimuth sectors.

[0112] In the embodiments of this application, the directional sector refers to the fan-shaped area formed by uniformly dividing the brush surface according to the angle, which is used to characterize the spatial distribution characteristics of different positions on the brush surface. The division method helps to analyze the development trend of temperature anomalies in different directions.

[0113] Please refer to Figure 3 , Figure 3 This is a schematic diagram of a directional sector division provided in an embodiment of this application.

[0114] Specifically, first, a polar coordinate system is established on the brush surface, with the brush center as the origin, and the entire brush surface is divided into multiple directional sectors at preset angular intervals (e.g., ...). Figure 3 (Sectors 1 to 8 in the data). For each identified anomalous region, the target orientation sector of anomalous region A is determined as sector 1 by calculating the centroid coordinates or boundary point distribution of the anomalous region (e.g., the centroid of anomalous region A is located in sector 1). When an anomalous region spans multiple sectors (e.g., anomalous region B spans sectors 3 and 7), the anomalous region can be assigned to the main covered sector based on its area proportion (e.g., anomalous region B covers a larger area in sector 3, so based on its area proportion, the target orientation sector of anomalous region B is sector 3).

[0115] S404: Calculate the dispersion of the temperature decay rate of the abnormal region in each target azimuth sector, and determine the target azimuth sector with the largest dispersion as the abnormal propagation region.

[0116] In the embodiments of this application, discreteness refers to the variance of the temperature decay rate of each abnormal region within the same sector, which is used to characterize the degree to which the temperature decay rate deviates from the average value.

[0117] Specifically, the average temperature decay rate within each sector is first calculated. Then, the difference between each temperature decay rate and the average is squared, and the sum of all squared terms is divided by the number of anomalous regions minus one. This yields the dispersion of the temperature decay rate within that sector. By comparing the variance of sectors in each target direction, the sector with the largest variance is identified as the anomalous propagation region. The underlying principle is that healthy brush materials are homogeneous, and their thermal conductivity should be roughly the same in all directions, resulting in a relatively concentrated temperature decay rate. When the temperature decay rate within a sector exhibits significant dispersion, it indicates a significant inhomogeneity in the material properties along that direction. For example, there may be internal microcracks extending along that direction or qualitative changes due to material ablation. These defects can hinder or accelerate heat conduction, leading to disordered temperature decay. Therefore, this region is the key area where anomalies are most likely to occur, develop, and propagate.

[0118] S405: Adjust the initial anomaly value based on the temperature decay rate in the anomaly propagation region to obtain the target anomaly value.

[0119] Specifically, the anomalous regions within the anomaly propagation area are first sorted according to their spatial distance from the dominant anomalous region, and the temperature decay rate difference between adjacent anomalous regions is calculated. By analyzing the changing trend of the temperature decay rate difference, the location where the difference changes abruptly is identified, and this location is determined as the temperature decay inflection point. Anomalous regions before the inflection point, due to their relatively stable temperature decay characteristics, are classified into the core anomalous region set.

[0120] Within the core anomaly region set, the arithmetic mean of the temperature decay rate of all anomaly regions is calculated. The arithmetic mean is then divided by the target temperature value of the dominant anomaly region to obtain the anomaly intensity coefficient, which reflects the degree of temperature anomaly.

[0121] Finally, the initial outlier is divided by the outlier intensity coefficient to obtain the target outlier after temperature feature correction.

[0122] Based on the above embodiments, as an optional embodiment, S405: the step of adjusting the initial anomaly value based on the temperature decay rate on the anomaly propagation region to obtain the target anomaly value may specifically include the following steps:

[0123] S501: Calculate the temperature decay rate difference between adjacent abnormal regions in the abnormal propagation region, and determine the temperature decay inflection point based on each temperature decay rate difference. The abnormal regions before the temperature decay inflection point are determined as the core abnormal region set.

[0124] In this embodiment, the physical meaning of the temperature decay inflection point can be understood as the core boundary of the heat-affected zone. Before the inflection point, the temperature decay rate changes gradually, indicating that these regions have a close heat exchange relationship with the dominant anomaly region (heat source), constituting the core of the anomaly. After the inflection point, the temperature decay rate changes sharply, indicating that the direct influence of the heat source on these regions weakens, belonging to the 'edge' or 'diffusion' region of the anomaly. By identifying the core region, the fundamental severity of the anomaly can be assessed more accurately, eliminating interference from peripheral secondary factors.

[0125] Specifically, firstly, within the anomalous propagation area, the anomalous areas are sorted according to their spatial distance from the dominant anomalous area to establish a spatial sequence. For adjacent anomalous areas, the difference in temperature decay rate is calculated, resulting in a series of temperature decay rate differences. This series of temperature decay rate differences is then fitted into a continuous curve using methods such as cubic spline interpolation. The first derivative of the fitted temperature decay rate curve is calculated to obtain the rate of change of the temperature decay rate; subsequently, the second derivative curve is obtained by differentiating the first derivative curve. By analyzing the sign change of the second derivative, the location where the second derivative changes from positive to negative, or from negative to positive, is the temperature decay inflection point. The anomalous areas spatially located before the temperature decay inflection point are defined as the core anomalous area set.

[0126] S502: Calculate the ratio of the average temperature decay rate within the core anomaly region set to the target temperature value of the dominant anomaly region to obtain the anomaly intensity coefficient.

[0127] In this embodiment, the abnormal intensity coefficient is a dimensionless index characterizing the spatial propagation characteristics and intensity distribution of abnormal brush temperature. From a physical perspective, it reflects the attenuation law of heat from the abnormal heat source during spatial propagation. When the abnormal intensity coefficient is close to 1, it indicates that the temperature attenuation is small as it propagates from the heat source to the surrounding area, suggesting severe heat accumulation and abnormal heat dissipation in the abnormal region. When the abnormal intensity coefficient is close to 0, it indicates that the temperature attenuation is significant as it propagates from the heat source to the surrounding area, suggesting normal heat dissipation in the abnormal region.

[0128] Specifically, the average temperature decay rate of all anomalous regions within the core anomalous region set is first calculated. This average reflects the overall heat conduction characteristics of the entire anomalous region. Then, the target temperature value of the dominant anomalous region is obtained, representing the source intensity of the anomalous heat generation. Dividing the average temperature decay rate of the core anomalous region set by the target temperature value of the dominant anomalous region yields the anomalous intensity coefficient. This calculation method, by comparing the overall temperature decay performance with the heat source intensity, considers both the spatial distribution characteristics of the anomalous region and its intensity level.

[0129] S503: Multiply the initial outlier by the reciprocal of the outlier intensity coefficient to obtain the target outlier.

[0130] In this embodiment, the target outlier refers to the temperature outlier after propagation characteristic compensation, which is used to characterize the true degree of temperature anomaly on the brush surface.

[0131] Specifically, the first step is to obtain the anomaly intensity coefficient, which reflects the characteristics of temperature propagation. Since heat gradually attenuates during propagation, the anomaly intensity coefficient is always less than 1. The reciprocal of the anomaly intensity coefficient is calculated to compensate for the attenuation effect during propagation. The detected initial anomaly value is multiplied by the reciprocal of the anomaly intensity coefficient to obtain the target anomaly value after eliminating the propagation effect. This correction method based on propagation characteristics, by compensating for the attenuation effect during heat transfer, can effectively eliminate the influence of uneven temperature propagation on anomaly detection and improve the accuracy of anomaly diagnosis. The adjustment logic here is that the 'anomaly intensity coefficient' characterizes the ease with which heat diffuses from the core area outwards. The smaller the value, the easier it is for heat to diffuse, the better the heat dissipation, and the relatively lower the actual hazard of the anomaly. Therefore, by multiplying it by its reciprocal (a number greater than 1), the 'initial anomaly value' can be amplified. Conversely, the larger the anomaly intensity coefficient, the more severe the heat accumulation, the poor the heat dissipation, and the higher the actual hazard; its reciprocal is close to 1, and the amplification effect on the initial anomaly value is weak.

[0132] S105: When the target abnormal value is higher than the preset abnormal threshold, it is determined that the brush has become abnormal.

[0133] Specifically, the target abnormal value is first obtained and compared with a preset abnormal threshold. When the target abnormal value exceeds the preset abnormal threshold, it is determined that the brush has malfunctioned.

[0134] Based on the above embodiments, as an optional embodiment, S105: after determining that the brush is abnormal, the step of replacing the brush is further included, which may specifically include the following steps:

[0135] S601: Based on visible light images and thermal infrared images, construct temporal and spatial data, which are data on the anomalies of the brush surface in time and space.

[0136] Specifically, visible light images provide morphological features of the brush surface, including visual characteristics such as wear level and surface defects; thermal infrared images provide temperature distribution information, reflecting the thermal anomalies of the brush. These two types of images are paired chronologically, and the location and size of abnormal regions in each pair are extracted to obtain spatial data between the visible light and thermal infrared images. For the spatial data at each time point, a timestamp is recorded to obtain spatiotemporal data.

[0137] S602: Based on time-space data, calculate the evolution rate of the anomalous region and the fluctuation frequency of the target temperature value.

[0138] In the embodiments of this application, the evolution rate refers to the speed at which the anomalous region expands and changes in the temporal and spatial dimensions.

[0139] Specifically, the process begins by extracting information on anomalous regions from the spatiotemporal data over continuous time series. The area of ​​the anomalous region at each time point is then calculated. The evolution rate is obtained by calculating the ratio of the change in anomalous region area between adjacent time points to the time interval. Simultaneously, a time-series analysis is performed on the target temperature value, counting the number of times the temperature value exceeds a preset threshold per unit time, thus obtaining the temperature fluctuation frequency.

[0140] S603: The product of the evolution rate and the fluctuation frequency is determined as the dynamic anomaly index.

[0141] Specifically, the evolution rate of the calculated anomalous region and the fluctuation frequency of the target temperature value are first obtained. The evolution rate is multiplied by the fluctuation frequency to obtain the dynamic anomaly index. This product combination includes both the development trend of the anomalous region in the spatial dimension and reflects the stability of temperature changes in the temporal dimension, forming a comprehensive spatiotemporal assessment system.

[0142] S604: Determine the dynamic abnormality index sequence within a preset abnormality time period after the brush malfunctions, and calculate the remaining service life of the brush based on the dynamic abnormality index sequence.

[0143] In the embodiments of this application, the remaining service life refers to the remaining operating time of the brush from the current moment to the moment when it needs to be replaced, which is used to characterize the length of time the brush can continue to operate safely.

[0144] Specifically, firstly, a dynamic anomaly index sequence is obtained within a preset abnormal time period, and the arithmetic mean of all dynamic anomaly indices in the sequence is calculated. The times corresponding to dynamic anomaly indices exceeding the mean are marked as high-risk times. The absolute value of the difference between the dynamic anomaly indices of adjacent high-risk times is calculated to obtain the fluctuation amplitude reflecting the severity of the anomaly. A preset fluctuation threshold is set as the judgment criterion, and the cumulative duration of high-risk time periods with fluctuation amplitudes exceeding the preset fluctuation threshold is counted to obtain the total risk duration. The total risk duration is divided by the total duration of all high-risk time periods to obtain the lifespan coefficient. Finally, the preset standard lifespan of the brush is multiplied by the lifespan coefficient to obtain the remaining service life.

[0145] Based on the above embodiments, as an optional embodiment, S604: the step of calculating the remaining service life of the brush based on the dynamic anomaly index sequence may specifically include the following steps:

[0146] S701: Calculate the mean of each dynamic anomaly index in the dynamic anomaly index sequence.

[0147] Specifically, firstly, a dynamic anomaly index sequence within a preset abnormal time period is obtained. This sequence contains dynamic anomaly index values ​​at consecutive time points. Then, the arithmetic mean of all dynamic anomaly index values ​​in the sequence is calculated to obtain the mean of the dynamic anomaly index.

[0148] S702: The time corresponding to the dynamic anomaly index that is greater than the mean is determined as a high-risk time. The absolute value of the difference in the dynamic anomaly index between each high-risk time is calculated to obtain the fluctuation range of the dynamic anomaly index.

[0149] Specifically, the time points corresponding to values ​​greater than the mean in the dynamic anomaly index sequence are first marked as high-risk moments. For adjacent high-risk moments, the absolute value of the difference in the dynamic anomaly index corresponding to the adjacent high-risk moments is calculated to obtain the fluctuation range. Using absolute value calculation can eliminate the directional influence and only focus on the magnitude of change.

[0150] S703: Calculate the total duration of high-risk periods with fluctuations exceeding a preset fluctuation threshold, and determine the lifespan coefficient as the ratio of the total duration of high-risk periods to the sum of all high-risk periods.

[0151] In this embodiment, the lifespan factor is a dimensionless index characterizing the degree of degradation of the brush under high-risk operating conditions. From a physical perspective, it reflects the cumulative effect of abnormal shocks suffered by the brush during high-risk periods: when the lifespan factor is close to 1, it indicates that most of the high-risk periods are accompanied by abnormal fluctuations, meaning that the brush is continuously in an unfavorable operating state, and its lifespan is accelerated.

[0152] Specifically, the calculated fluctuation amplitude sequence is first obtained, and each fluctuation amplitude is compared with a preset fluctuation threshold. For high-risk moments where the fluctuation amplitude exceeds the preset threshold, the time intervals corresponding to two adjacent high-risk moments are counted and summed to obtain the total risk duration. Simultaneously, the total length of all high-risk time intervals is calculated, including those where the fluctuation amplitude does not exceed the preset threshold. The total risk duration is then divided by the sum of all high-risk time intervals to obtain the lifespan coefficient.

[0153] S704: Multiply the preset standard lifespan of the brush by the lifespan coefficient to obtain the remaining lifespan of the brush.

[0154] Specifically, first, obtain the preset standard lifespan of the brush, then multiply the preset standard lifespan by the calculated lifespan coefficient to obtain the remaining lifespan.

[0155] S605: When the trend of the dynamic abnormal index meets the preset warning conditions, calculate the optimal brush replacement time window based on the remaining service life.

[0156] In this embodiment, the changing trend of the dynamic anomaly index represents the growth rate of the dynamic anomaly index, and the warning condition refers to the state in which the growth rate of the anomaly index exceeds a preset threshold.

[0157] Specifically, by monitoring the changes in the dynamic anomaly index in real time, when the growth rate of the dynamic anomaly index meets the preset warning conditions, a certain maintenance preparation time is reserved in advance as the start time of the time window, with the remaining service life as the baseline time point. The maintenance preparation time must be sufficient to complete the preparation of spare parts and personnel allocation. The safety margin time before the expiration of the remaining service life is used as the end time of the time window. The safety margin time is the necessary maintenance execution time, and finally the optimal brush replacement time window is determined.

[0158] S606: Within the optimal brushing time window, select the idle time point of the train as the target brushing time according to the train operation plan.

[0159] Specifically, within the optimal brush replacement time window, time information from the train operation plan is obtained, including departure time, arrival time, and turnaround time. By analyzing the operation plan, various idle periods of the train within the optimal brush replacement time window are identified. The duration of each idle period is evaluated to ensure that the minimum time requirement for brush replacement is met. Among the idle periods that meet the duration requirement, the start time of the longest period is selected as the target brush replacement time.

[0160] refer to Figure 4 This application also provides a schematic diagram of a contact current collector brush monitoring system, which specifically includes:

[0161] The image acquisition module is used to acquire a visible light image and a thermal infrared image of the brush at the same location when the brush is in contact with the contact wire and carrying the working current while the train is running.

[0162] The initial outlier calculation module is used to identify the geometric contour of the brush based on the visible light image, calculate the current remaining thickness of the brush in the direction perpendicular to the contact wire based on the geometric contour, and calculate the initial outlier based on the current remaining thickness.

[0163] The target temperature value extraction module is used to determine the current exchange zone based on the visible light image. The current exchange zone is the continuous area where the brush and the contact wire are pressed and contacted at the current moment. Based on the thermal infrared image, the surface of the brush is divided into multiple sub-regions, the target sub-region located in the current exchange zone is selected, and the target temperature value of the target sub-region is extracted.

[0164] The target anomaly calculation module is used to identify target sub-regions whose target temperature values ​​are higher than the average temperature value of the current exchange zone as anomaly regions, and to add the areas of each of the anomaly regions to obtain the cumulative anomaly area. The initial anomaly value is adjusted according to the cumulative anomaly area to obtain the target anomaly value.

[0165] The anomaly determination module is used to determine that the brush has malfunctioned when the target anomaly value is higher than a preset anomaly threshold.

[0166] Optionally, the initial outlier calculation module is specifically used for:

[0167] Obtain the first remaining thickness sequence of the brush within a first preset time period before the current time, divide the first remaining thickness sequence into multiple time windows, and calculate the thickness change rate of the brush within each time window;

[0168] Based on the thickness change rate and the current remaining thickness, determine the first thickness change value of the brush within a second preset time period after the current moment;

[0169] Obtain the intermediate thickness sequence of the brush within a third preset time period before the current time. Based on the intermediate thickness sequence and the first thickness change value, predict the second thickness change value of the brush within a fourth preset time period after the current time. The third preset time period is greater than and includes the first preset time period, and the fourth preset time period is greater than and includes the second preset time period.

[0170] Calculate the weighted average of the first thickness change value and the second thickness change value, and divide the difference between the current remaining thickness and the weighted average by the weighted average to obtain the initial outlier value.

[0171] Optionally, the initial outlier calculation module is further specifically used for:

[0172] Calculate the difference between two adjacent thickness values ​​in the intermediate thickness sequence, and form a difference sequence from each difference. Calculate the wear rate of the brush at each moment based on the difference sequence.

[0173] The moment when the wear rate is greater than the first rate threshold is defined as the abnormal wear moment, and the cumulative wear value within a preset wear time range before and after each abnormal wear moment is calculated;

[0174] The difference between adjacent cumulative wear values ​​is calculated, and the average of the differences between the cumulative wear values ​​is subtracted from the first thickness change value to obtain the second thickness change value of the brush within the fourth preset time period after the current moment.

[0175] Optionally, the target outlier calculation module is specifically used for:

[0176] The area weighting coefficient of each abnormal region is obtained based on the proportion of the area of ​​each abnormal region to the total area of ​​abnormalities.

[0177] The abnormal region with the highest target temperature value is identified as the dominant abnormal region. Based on the area weighting coefficient of each abnormal region, the temperature decay rate of each abnormal region relative to the dominant abnormal region is calculated.

[0178] The target azimuth sector where each of the abnormal regions is located is determined, and the brush is composed of multiple preset azimuth sectors;

[0179] Calculate the dispersion of the temperature decay rate of the abnormal region in each of the target azimuth sectors, and determine the target azimuth sector with the largest dispersion as the abnormal propagation region;

[0180] The initial anomaly value is adjusted based on the temperature decay rate in the anomaly propagation region to obtain the target anomaly value.

[0181] Optionally, the target outlier calculation module is further specifically used for:

[0182] Calculate the temperature decay rate difference between adjacent abnormal regions in the abnormal propagation region, and determine the temperature decay inflection point based on each temperature decay rate difference. The abnormal regions before the temperature decay inflection point are determined as the core abnormal region set.

[0183] The ratio of the average temperature decay rate within the core anomaly region set to the target temperature value of the dominant anomaly region is calculated to obtain the anomaly intensity coefficient.

[0184] The target anomaly value is obtained by multiplying the initial anomaly value by the reciprocal of the anomaly intensity coefficient.

[0185] Optionally, a contact current collector brush monitoring system further includes a brush replacement module, specifically used for:

[0186] Based on the visible light image and the thermal infrared image, temporal and spatial data are constructed, which are data on the anomalies of the brush surface in time and space.

[0187] Based on the spatiotemporal data, the evolution rate of the abnormal region and the fluctuation frequency of the target temperature value are calculated.

[0188] The product of the evolution rate and the fluctuation frequency is determined as the dynamic anomaly index;

[0189] Determine the dynamic anomaly index sequence within a preset anomaly time period after the brush malfunctions, and calculate the remaining service life of the brush based on the dynamic anomaly index sequence.

[0190] When the changing trend of the dynamic anomaly index meets the preset early warning conditions, the optimal brush replacement time window is calculated based on the remaining service life.

[0191] Within the optimal brushing time window, the idle time point of the train is selected as the target brushing time according to the train operation plan.

[0192] Optionally, the brush replacement module is further specifically used for:

[0193] Calculate the mean of each dynamic anomaly index in the dynamic anomaly index sequence;

[0194] The time corresponding to the dynamic anomaly index that is greater than the mean is determined as a high-risk time. The absolute value of the difference of the dynamic anomaly index between each high-risk time is calculated to obtain the fluctuation range of the dynamic anomaly index.

[0195] The total duration of high-risk periods with fluctuations exceeding a preset fluctuation threshold is calculated, and the ratio of this total duration to the sum of all high-risk periods is determined as the lifespan coefficient.

[0196] The remaining service life of the brush is obtained by multiplying the preset standard lifespan of the brush by the lifespan coefficient.

[0197] It should be noted that the above embodiments of the apparatus are only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0198] This embodiment also discloses an electronic device, referring to... Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. The electronic device 013 may include: at least one processor 901, at least one communication bus 902, a user interface 903, a network interface 904, and at least one memory 905.

[0199] The communication bus 902 is used to enable communication between these components.

[0200] The user interface 903 may include a display screen and a camera. Optionally, the user interface 903 may also include a standard wired interface and a wireless interface.

[0201] The network interface 904 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0202] The processor 901 may include one or more processing cores. The processor 901 connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 905, and by calling data stored in the memory 905. Optionally, the processor 901 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array. The processor 901 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content to be displayed on the screen; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 901 and may be implemented as a separate chip.

[0203] The memory 905 may include random access memory (RAM) or read-only memory. Optionally, the memory 905 may include a non-transitory computer-readable storage medium. The memory 905 may be used to store instructions, programs, code, code sets, or instruction sets. The memory 905 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 905 may also be at least one storage device located remotely from the aforementioned processor 901. (See reference...) Figure 5 The memory 905, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for monitoring the brushes of a contact current collector.

[0204] exist Figure 5In the electronic device shown, the user interface 903 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 901 can be used to call an application program for monitoring the contact current collector brush stored in the memory 905. When executed by one or more processors 901, the electronic device 013 performs one or more methods as described in the above embodiments.

[0205] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0206] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0207] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings or direct couplings or communication connections may be through some service interfaces; indirect couplings or communication connections between apparatuses or units may be electrical or other forms.

[0208] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0209] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0210] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0211] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Other embodiments of this disclosure will be readily apparent to those skilled in the art upon consideration of the disclosure in this specification. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method for monitoring the brush of a contact current collector, characterized in that, Applied to electronic devices, the method includes: When the train is in operation, and the brush is in contact with the contact wire and carrying the working current, a visible light image and a thermal infrared image of the brush are acquired at the same location at the current moment. Based on the visible light image, the geometric contour of the brush is identified, and the current remaining thickness of the brush in the direction perpendicular to the contact wire is calculated based on the geometric contour. An initial outlier value is then calculated based on the current remaining thickness. The calculation of the initial outlier value based on the current remaining thickness includes: Obtain the first remaining thickness sequence of the brush within a first preset time period before the current time, divide the first remaining thickness sequence into multiple time windows, and calculate the thickness change rate of the brush within each time window; Based on the thickness change rate and the current remaining thickness, determine the first thickness change value of the brush within a second preset time period after the current moment; Obtain the intermediate thickness sequence of the brush within a third preset time period before the current time. Based on the intermediate thickness sequence and the first thickness change value, predict the second thickness change value of the brush within a fourth preset time period after the current time. The third preset time period is greater than and includes the first preset time period, and the fourth preset time period is greater than and includes the second preset time period. Calculate the weighted average of the first thickness change value and the second thickness change value, and divide the difference between the current remaining thickness and the weighted average by the weighted average to obtain the initial outlier value; Based on the visible light image, the current exchange zone is determined. The current exchange zone is the continuous area where the brush and the contact wire are in contact at the current moment. Based on the thermal infrared image, the surface of the brush is divided into multiple sub-regions. The target sub-region located in the current exchange zone is selected, and the target temperature value of the target sub-region is extracted. The target sub-regions whose target temperature values ​​are higher than the average temperature value of the current exchange region are identified as abnormal regions. The areas of the abnormal regions are added together to obtain the cumulative abnormal area. The initial abnormal value is adjusted according to the cumulative abnormal area to obtain the target abnormal value. The step of adjusting the initial outlier value based on the cumulative area of ​​the outlier to obtain the target outlier value includes: The area weighting coefficient of each abnormal region is obtained based on the proportion of the area of ​​each abnormal region to the total area of ​​abnormalities. The abnormal region with the highest target temperature value is identified as the dominant abnormal region. Based on the area weighting coefficient of each abnormal region, the temperature decay rate of each abnormal region relative to the dominant abnormal region is calculated. The target azimuth sector where each of the abnormal regions is located is determined, and the brush is composed of multiple preset azimuth sectors; Calculate the dispersion of the temperature decay rate of the abnormal region in each of the target azimuth sectors, and determine the target azimuth sector with the largest dispersion as the abnormal propagation region; The initial anomaly value is adjusted based on the temperature decay rate in the anomaly propagation region to obtain the target anomaly value; When the target abnormal value is higher than the preset abnormal threshold, it is determined that the brush has malfunctioned.

2. The method for monitoring the brush of a contact current collector according to claim 1, characterized in that, The step of predicting the second thickness change value of the brush within a fourth preset time period after the current moment based on the intermediate thickness sequence and the first thickness change value includes: Calculate the difference between two adjacent thickness values ​​in the intermediate thickness sequence, and form a difference sequence from each difference. Calculate the wear rate of the brush at each moment based on the difference sequence. The moment when the wear rate is greater than the first rate threshold is defined as the abnormal wear moment, and the cumulative wear value within a preset wear time range before and after each abnormal wear moment is calculated; The difference between adjacent cumulative wear values ​​is calculated, and the average of the differences between the cumulative wear values ​​is subtracted from the first thickness change value to obtain the second thickness change value of the brush within the fourth preset time period after the current moment.

3. The method for monitoring the brush of a contact current collector according to claim 1, characterized in that, The adjustment of the initial anomaly value based on the temperature decay rate in the anomaly propagation region to obtain the target anomaly value includes: Calculate the temperature decay rate difference between adjacent abnormal regions in the abnormal propagation region, and determine the temperature decay inflection point based on each temperature decay rate difference. The abnormal regions before the temperature decay inflection point are determined as the core abnormal region set. The ratio of the average temperature decay rate within the core anomaly region set to the target temperature value of the dominant anomaly region is calculated to obtain the anomaly intensity coefficient. The target anomaly value is obtained by multiplying the initial anomaly value by the reciprocal of the anomaly intensity coefficient.

4. The method for monitoring the brush of a contact current collector according to claim 1, characterized in that, After determining that the brush has malfunctioned, the process further includes: Based on the visible light image and the thermal infrared image, temporal and spatial data are constructed, which are data on the anomalies of the brush surface in time and space. Based on the spatiotemporal data, the evolution rate of the abnormal region and the fluctuation frequency of the target temperature value are calculated. The product of the evolution rate and the fluctuation frequency is determined as the dynamic anomaly index; Determine the dynamic anomaly index sequence within a preset anomaly time period after the brush malfunctions, and calculate the remaining service life of the brush based on the dynamic anomaly index sequence. When the changing trend of the dynamic anomaly index meets the preset early warning conditions, the optimal brush replacement time window is calculated based on the remaining service life. Within the optimal brushing time window, the idle time point of the train is selected as the target brushing time according to the train operation plan.

5. The method for monitoring the brush of a contact current collector according to claim 4, characterized in that, The calculation of the remaining service life of the brush based on the dynamic anomaly index sequence includes: Calculate the mean of each dynamic anomaly index in the dynamic anomaly index sequence; The time corresponding to the dynamic anomaly index that is greater than the mean is determined as a high-risk time. The absolute value of the difference of the dynamic anomaly index between each high-risk time is calculated to obtain the fluctuation range of the dynamic anomaly index. The total duration of high-risk periods with fluctuations exceeding a preset fluctuation threshold is calculated, and the ratio of this total duration to the sum of all high-risk periods is determined as the lifespan coefficient. The remaining service life of the brush is obtained by multiplying the preset standard lifespan of the brush by the lifespan coefficient.

6. A contact current collector brush monitoring system, applied to the contact current collector brush monitoring method as described in any one of claims 1-5, characterized in that, The system is applied to electronic devices and includes: The image acquisition module is used to acquire a visible light image and a thermal infrared image of the brush at the same location when the brush is in contact with the contact wire and carrying the working current while the train is running. The initial outlier calculation module is used to identify the geometric contour of the brush based on the visible light image, calculate the current remaining thickness of the brush in the direction perpendicular to the contact wire based on the geometric contour, and calculate the initial outlier based on the current remaining thickness. The target temperature value extraction module is used to determine the current exchange zone based on the visible light image. The current exchange zone is the continuous area where the brush and the contact wire are in contact at the current moment. Based on the thermal infrared image, the surface of the brush is divided into multiple sub-regions, the target sub-region located in the current exchange zone is selected, and the target temperature value of the target sub-region is extracted. The target anomaly calculation module is used to identify target sub-regions whose target temperature values ​​are higher than the average temperature value of the current exchange zone as anomaly regions, and to add the areas of each of the anomaly regions to obtain the cumulative anomaly area. The initial anomaly value is adjusted according to the cumulative anomaly area to obtain the target anomaly value. The anomaly determination module is used to determine that the brush has malfunctioned when the target anomaly value is higher than a preset anomaly threshold.

7. An electronic device, characterized in that, The device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions. The user interface and the network interface are both used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed, perform the method as described in any one of claims 1-5.

Images