Railway contact network state prediction method and related system

Abstract

The invention provides a railway overhead line system state prediction method and a related system, and belongs to the related technical field of electrified railways. Specifically, meteorological data and equipment multi-dimensional monitoring data of a railway contact network site are collected in real time; inputting the acquired meteorological data and the equipment multi-dimensional monitoring data into a pre-trained prediction model, and performing equipment state trend analysis and fault early warning according to a prediction result; and the prediction model adopts a multi-target regression analysis model to perform fault prediction. By constructing and applying prediction models such as linear regression, historical and real-time data are converted into predicted values of future equipment state parameters. Therefore, the basis of operation and maintenance decision is changed from occurred abnormities to occurred risks. According to the method, the parameter degradation trend can be identified in advance, and the early warning is given out before the actual out-of-limit of the equipment, so that the unplanned shutdown is effectively avoided, and the accident risk and the maintenance cost are reduced.

Description

Technical Field

[0001] This invention belongs to the technical field of electrified railways, specifically relating to a method for predicting the status of railway catenary and related systems. Background Technology

[0002] The railway overhead contact system is a core component of the electrified railway traction power supply system. Structurally, it is erected above the track in a zigzag pattern and mainly consists of contact wires, catenary wires, support devices, and supports. The primary function of this system is to continuously and stably transmit electrical energy to the pantograph on top of moving electric locomotives or EMUs via sliding contact. To ensure the reliability of power supply and the safety of train operation, the design and construction of the overhead contact system must comprehensively consider various complex climatic conditions, including wind, rain, snow, ice, and temperature changes, and be able to withstand the dynamic mechanical loads generated by high-speed train operation, arcing impacts, and long-term mechanical wear to ensure that the contact wire and pantograph maintain good electrical and mechanical contact at all times.

[0003] Given that the overhead contact line is the only direct interface for railway traction power supply, its operational integrity directly affects the continuity and stability of train traction power supply, and even the operational safety and efficiency of the entire railway transportation system. If the overhead contact line experiences problems such as wire breakage, detachment, excessive wear, component failure, or exceeding geometric parameters (such as conductor height and pull-out value), it can lead to anything from minor train power outages and operational interruptions to serious accidents, causing significant loss of life, property damage, and economic losses.

[0004] Currently, the monitoring and maintenance of the overhead contact system in my country's railways mainly relies on the traditional, periodic manual inspection model. This model typically involves professionals conducting inspections by visual observation and simple measuring tools through ground patrols, locomotive or work vehicle inspections. This model has significant limitations: First, it is inefficient, with limited coverage and long inspection cycles, making it difficult to achieve real-time coverage across the entire line and prone to blind spots. Second, it is slow to respond, making it difficult to detect and warn of sudden faults or hidden dangers in a timely manner. There is often a significant time delay between the occurrence and discovery of a fault, which cannot meet the stringent requirements of high-density, high-speed operation for safety response time in high-speed railways. Third, it is greatly affected by subjective and environmental factors, with inspection quality depending on the experience and sense of responsibility of the personnel, and the inspection effect is greatly reduced in adverse weather or at night. In addition, there is a lack of precise control over dynamic operating parameters, making it difficult for manual static inspection to effectively assess the dynamic performance of the overhead contact system when a train's pantograph passes at high speed.

[0005] With the rapid development of electrified railways, especially high-speed railways, and the continuous increase in operating mileage, train speeds and frequencies are constantly increasing, placing higher demands on the safe and reliable operation of the overhead contact system. The shortcomings of traditional manual inspection methods in terms of monitoring efficiency, real-time performance, accuracy, and automation are becoming increasingly apparent, posing a bottleneck to further improving railway operational safety. Therefore, there is an urgent need to develop and apply more advanced and intelligent real-time monitoring technologies and systems for the overhead contact system to achieve automated, continuous, and high-precision monitoring and intelligent diagnosis of key parameters and conditions, transforming periodic maintenance into predictive maintenance, thereby significantly improving the safety level, operational efficiency, and economic benefits of the railway power supply system. Summary of the Invention

[0006] The purpose of this invention is to overcome the above-mentioned shortcomings and provide a method and related system for predicting the condition of railway catenary.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for predicting the condition of a railway catenary, comprising the following steps: Real-time collection of meteorological data and multi-dimensional monitoring data of equipment at the railway catenary site; The collected meteorological data and multi-dimensional monitoring data of the equipment are input into a pre-trained prediction model, the prediction results are output, and the equipment status trend analysis and fault warning are performed based on the prediction results. The prediction model is built based on historical monitoring data and uses a multi-objective regression analysis model for fault prediction.

[0008] In the step of real-time acquisition of meteorological data and multi-dimensional monitoring data of the railway catenary, the meteorological data includes wind speed, wind direction, temperature, humidity and atmospheric pressure, and the multi-dimensional monitoring data of the equipment includes basic information of the equipment, anchor offset of the catenary, tension, clamp temperature, insulator leakage current, static angle and dynamic amplitude.

[0009] The method for constructing and training the prediction model in the steps of inputting the collected meteorological data and multi-dimensional monitoring data of the equipment into a pre-trained prediction model, outputting prediction results, and performing equipment status trend analysis and fault early warning based on the prediction results is as follows: A historical dataset is acquired, and the data in the historical dataset is cleaned by removing missing values ​​to obtain a preprocessed historical dataset; the historical data includes historical meteorological data and multi-dimensional monitoring data from equipment. The preprocessed historical dataset is divided into feature variables and target variables; The feature variables and target variables are divided into training sets and test sets, a multi-objective regression analysis model is initialized, and the initialized multi-objective regression analysis model is trained using the training set to generate an initial prediction model; The initial prediction model is used to make predictions using the test set to obtain prediction results. The initial prediction model is then trained based on the residuals between the prediction results and the true values ​​to obtain a trained prediction model.

[0010] The multi-objective regression analysis model adopts a multi-output linear regression model, a multi-task learning model, a decision tree-based multi-output regression model, or a deep neural network model.

[0011] The method for dividing the preprocessed historical dataset into feature variables and target variables is as follows: Based on the preprocessed historical dataset, target variables corresponding to the prediction target are selected, and a distribution histogram of the target variables is plotted to analyze their overall distribution characteristics. The other columns of data in the preprocessed dataset, excluding the target variable, are defined as the feature variable set, and the target variable is defined as the prediction label.

[0012] The method for training the initial prediction model in the step of using a test set to predict the initial prediction model, obtaining the prediction result, and training the initial prediction model based on the residual between the prediction result and the true value to obtain a trained prediction model is as follows: Calculate the mean squared error and coefficient of determination between the predicted results and the actual values ​​to quantitatively evaluate the prediction accuracy and interpretability of the initial prediction model under the default hyperparameters. The residual between the predicted result and the actual value is calculated to obtain the residual calculation result. Based on the residual calculation result, a residual distribution histogram is plotted to test whether the error meets the basic assumptions of linear regression, thereby diagnosing whether there is a systematic bias in the initial prediction model. If the initial prediction model is found to have systematic biases, a grid search method is used to traverse the preset hyperparameter space, with the goal of minimizing the mean squared error. The optimal combination of hyperparameters is selected through cross-validation, thereby obtaining a well-trained prediction model.

[0013] The method for analyzing equipment status trends and providing fault warnings based on prediction results is as follows: the prediction results are compared with preset alarm threshold parameters. If the predicted value or the real-time value of the equipment's multi-dimensional monitoring data exceeds the preset alarm threshold, an alarm is triggered and alarm information is generated.

[0014] Secondly, the present invention provides a railway catenary condition prediction system, comprising: The data acquisition module is used to collect meteorological data and multi-dimensional monitoring data of the railway catenary in real time. The prediction module is used to input the collected meteorological data and multi-dimensional monitoring data of the equipment into the pre-trained prediction model, output the prediction results, and perform equipment status trend analysis and fault early warning based on the prediction results. The prediction model is built based on historical monitoring data and uses a multi-objective regression analysis model for fault prediction.

[0015] Thirdly, the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a railway catenary state prediction method.

[0016] Fourthly, the present invention provides a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of a railway catenary state prediction method.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This invention provides a method for predicting the condition of railway overhead contact lines. By constructing and applying predictive models such as linear regression, historical and real-time data are transformed into predicted values ​​for future equipment condition parameters. This shifts the basis for maintenance decisions from past anomalies to anticipated risks. This method can identify parameter degradation trends in advance and issue warnings before equipment actually exceeds limits, effectively avoiding unplanned outages and reducing accident risks and maintenance costs. By constructing automated and intelligent monitoring and prediction methods, it effectively solves the core pain points of traditional manual inspection modes, such as low efficiency, poor real-time performance, low accuracy, and low automation. It transforms discrete, periodic manual inspections into continuous, comprehensive, data-driven monitoring, achieving an exponential increase in monitoring efficiency, a leap in real-time condition warnings, more objective and accurate fault diagnosis, and a high degree of automation in the maintenance process.

[0018] Furthermore, traditional methods often set independent alarms for individual parameters, lacking correlation analysis between alarms and making it difficult to assess overall risk. This invention employs a multi-output regression model to achieve parallel and synchronous prediction of multiple key parameters such as anchor offset, tension, and temperature. It overcomes the limitations of single-point alarms, enabling a comprehensive health assessment of the entire contact network subsystem (such as an anchor section) in a single, highly efficient manner, resulting in more comprehensive early warning information and more reliable decision support. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a system diagram of Embodiment 3 of the present invention. Detailed Implementation

[0020] To further understand the content of this invention, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments are merely illustrative and not limiting of the invention.

[0021] Example 1 like Figure 1 As shown, a method for predicting the condition of a railway catenary includes the following steps: S1: Real-time collection of meteorological data and multi-dimensional monitoring data of equipment at the railway catenary site; S2: Input the collected meteorological data and multi-dimensional monitoring data of the equipment into the pre-trained prediction model, output the prediction results, and perform equipment status trend analysis and fault warning based on the prediction results; The prediction model is built based on historical monitoring data and uses a multi-objective regression analysis model for fault prediction.

[0022] Specifically, in S1, meteorological data and multi-dimensional monitoring data of the railway catenary are collected in real time, and the data is transmitted to the data server via a wireless network.

[0023] Meteorological data includes wind speed, wind direction, temperature, humidity, and atmospheric pressure. Multi-dimensional equipment monitoring data includes basic equipment information, contact wire anchor offset, tension, clamp temperature, insulator leakage current, static angle, and dynamic amplitude. The collected real-time data is transmitted to the data server via a 4G network, ensuring fast and stable data transmission. After receiving this meteorological and multi-dimensional equipment monitoring data, the data server distributes it to the display platform in real time for users to view and analyze.

[0024] Meteorological data from the railway catenary is collected using five-element meteorological sensors and transmitted to a data aggregator via an RS485 interface.

[0025] The core objective of equipment operation monitoring is to track the operational status of equipment in real time to ensure its normal operation. This includes real-time monitoring of the equipment's online status. When the equipment is online, the system indicates it with a green indicator; when the equipment is offline, the system indicates it with a red indicator; and when the equipment is in a dormant state, the system indicates it with a gray indicator. These indicators help maintenance personnel quickly identify the current status of the equipment and take appropriate measures in a timely manner.

[0026] To better manage equipment, basic equipment information includes, but is not limited to, equipment serial number, equipment name, equipment installation location, platform marker, and equipment type. By configuring this information, maintenance personnel can more easily manage and maintain the equipment, ensuring its efficient operation.

[0027] Data aggregators play a crucial role in equipment operation monitoring. They are responsible for collecting real-time heartbeat data from various devices in the field, which is key to determining the equipment's operating status. Through analysis and processing of this data, the data aggregator can determine the equipment's operating status based on the data type, thus providing an accurate foundation for subsequent data processing and transmission.

[0028] Subsequently, the operational status information of these devices is transmitted to a real-time data server via the 4G network. Upon receiving this information, the real-time data server performs further processing and analysis to ensure the accuracy and timeliness of the information. Then, the real-time data server transmits the device operational status to the platform in real time, ensuring that the platform can obtain the latest device status information promptly. After receiving the device operational status information, the platform displays the devices in different formats on the simulated railway line map according to their operational status.

[0029] Preferably, the online monitoring of equipment uses charts to statistically analyze and display the number of online devices in real time, allowing users to intuitively grasp the real-time online status of the equipment.

[0030] Specifically, in S2, the collected meteorological data and multi-dimensional monitoring data of the equipment are input into the pre-trained prediction model, the prediction results are output, and the equipment status trend analysis and fault warning are performed based on the prediction results.

[0031] The prediction model is built based on historical monitoring data and uses a multi-objective regression analysis model for fault prediction. The construction and training methods of the prediction model are as follows: 1) Obtain historical datasets from the data server, clean the data in the historical datasets, remove missing values, and obtain preprocessed historical datasets; the historical datasets include historical meteorological data and multi-dimensional monitoring data from equipment; 2) Divide the preprocessed historical dataset into feature variables and target variables; Based on the preprocessed historical dataset, target variables corresponding to the prediction target are selected, and a distribution histogram of the target variables is plotted to analyze its overall distribution characteristics; the other columns of data in the preprocessed dataset excluding the target variable are defined as the feature variable set X, and the target variable is defined as the prediction label y; 3) Divide the feature variables and target variables into training sets and test sets, initialize a multi-objective regression analysis model, and train the initialized multi-objective regression analysis model using the training set to generate an initial prediction model; 4) Use the test set to make predictions on the initial prediction model, obtain the prediction results, and train the initial prediction model based on the residuals between the prediction results and the true values ​​to obtain a trained prediction model. The training method is as follows: The mean square error and coefficient of determination between the predicted results and the actual values ​​are calculated to quantitatively evaluate the prediction accuracy and interpretability of the initial prediction model under the default hyperparameters.

[0032] Furthermore, the residual between the predicted result and the true value is calculated to obtain the residual calculation result. Based on the residual calculation result, a residual distribution histogram is plotted to test whether the error conforms to the basic assumptions of linear regression, thereby diagnosing whether there is a systematic bias in the initial prediction model. If the diagnostic results indicate that the initial prediction model structure is applicable but its accuracy needs to be improved, a grid search method is used to traverse the preset hyperparameter space, with the mean square error in S4 as the optimization objective. The optimal combination of hyperparameters is selected through cross-validation, thereby training a better-performing optimized prediction model and obtaining a well-trained prediction model.

[0033] Preferably, the multi-objective regression analysis model can be any of the following models or variations thereof: multi-output linear regression model, multi-task learning model, decision tree-based multi-output regression model, or deep neural network model.

[0034] In this preferred embodiment, the dataset in CSV format is imported into a DataFrame object using the read_csv function from the pandas library. Then, the dropna() function is executed to remove rows containing missing values ​​to avoid errors during subsequent analysis.

[0035] Next, a histogram of the target variable (i.e., the target column) is plotted to explore its distribution characteristics. This helps to understand the overall distribution of the data, including potential skewness and outliers. All columns except the target variable are treated as features (x), while the target variable itself is used as the prediction target (y).

[0036] The `train_test_split` function splits the dataset into a training set (x_train, y_train) and a test set (X_test, y_test), with the test set comprising 20% ​​of the total data. The `random_state` parameter is set to ensure consistency in each split. Next, a linear regression model is created by instantiating the `LinearRegression` class, and the model is trained using the training set to generate an initial prediction model.

[0037] The test set is input into the initial prediction model to make predictions, and the prediction results are obtained. The mean squared error (MSE) and the coefficient of determination (R²) between the prediction results and the actual results are calculated to evaluate the performance of the model.

[0038] Furthermore, the residuals between the predicted and actual values ​​(i.e., the difference between the actual and predicted values) are calculated, and a residual distribution histogram is plotted. This helps diagnose whether the model has bias or variance issues; ideally, the residuals should exhibit an approximately normal distribution. Finally, the GridSearchCV method is used to optimize the model parameters to improve model performance, and the trained model is used to predict new data, outputting the model's hyperparameters. These hyperparameters reflect the degree of influence of each feature on the target variable.

[0039] Furthermore, the prediction results are compared with the alarm threshold parameters preset in the parameter configuration interface. If the predicted value or the real-time value of the device's multi-dimensional monitoring data exceeds the preset alarm threshold, an alarm is triggered and scrolling alarm information is generated in the device's alarm monitoring area.

[0040] Furthermore, it also includes a report generation tool that supports generating monitoring reports and exporting them in multiple formats, making it convenient for users to share and archive reports.

[0041] The preset alarm threshold parameters are as follows: The allowable range of the anchoring parameters (ranging A and ranging B) for the overhead contact line should exceed 200 mm. The tension parameter should allow a 15% fluctuation based on a 15 kN base. The temperature of the disconnecting switch and electrical connection clamps should not exceed 90°C. The insulator leakage current parameter should be between 30 μA and 100 mA. The parameters of the catenary wire, positioner, and combined static angle (zAngle) should be between 7 and 17 degrees. The parameters of the catenary wire and combined dynamic amplitude (z) should be between -100 mm and 100 mm.

[0042] The equipment operating parameters are specified as follows: Low power sleep parameters: Low power sleep duration is set to the default 24 hours; Low power sleep voltage is set to the default 2300 millivolts.

[0043] Scanning device parameters: The sleep time for other devices except the insulator collector is set to the default 5 minutes; the sleep time for the insulator collector is set to the default 60 minutes.

[0044] Synchronization device parameters: the acquisition frequency is set to the default 100 Hz; the threshold for the number of vibration peaks is set to the default 2; the threshold for the number of vibration data packets is set to the default 6 packets; the threshold for the number of stationary data packets is set to the default 60 packets; and the threshold for the number of empty data packets is set to the default 60 packets.

[0045] Example 2 A railway catenary condition prediction system, comprising: The data acquisition module is used to collect meteorological data and multi-dimensional monitoring data of the railway catenary in real time. The prediction module is used to input the collected meteorological data and multi-dimensional monitoring data of the equipment into the pre-trained prediction model, output the prediction results, and perform equipment status trend analysis and fault early warning based on the prediction results. The prediction model is built based on historical monitoring data and uses a multi-objective regression analysis model for fault prediction.

[0046] Furthermore, the interface includes a parameter configuration interface, i.e., an interactive interface, comprising: a meteorological monitoring data area, which displays real-time on-site weather conditions in the form of a container panel, providing intuitive data for meteorological analysis; an equipment operation status monitoring area, which displays the equipment installation location through on-site simulation and scans the operation status of each device in real-time using a carousel format, ensuring comprehensive control over the equipment status; an online equipment monitoring area, which uses a graphical method to display real-time statistics on the number of online devices, providing users with immediate online status information; and an equipment alarm monitoring area, which displays equipment alarm information in a seamless scrolling table format, ensuring that alarm information can be detected and handled promptly.

[0047] The parameter configuration interface includes a historical data query function, allowing users to review and analyze historical monitoring data for deeper insights. Simultaneously, the interface supports real-time data analysis and displays the results in charts for easy user understanding of data changes. This interface also allows users to adjust equipment alarm parameters and calibrate equipment operating parameters according to on-site needs, ensuring stable equipment operation and timely response.

[0048] By configuring alarm and equipment parameters in the parameter configuration interface, the platform can monitor and trigger alarms according to user needs. Real-time analysis of equipment online status, real-time monitoring of on-site weather, and equipment alarm information ensures a comprehensive understanding of all indicators. Real-time monitoring of equipment data and exporting monitoring reports and data charts according to user requirements provide strong support for decision-making.

[0049] Example 3 like Figure 2 As shown, the present invention also provides an electronic device 100 for a railway catenary condition prediction method; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0050] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the railway catenary state prediction method described in Embodiment 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0051] The at least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor. The processor 102 is the control center of the electronic device 100, connecting various parts of the electronic device 100 via various interfaces and lines.

[0052] The memory 101 in the electronic device 100 stores multiple instructions to implement a railway catenary state prediction method, and the processor 102 can execute the multiple instructions to achieve the following: Real-time collection of meteorological data and multi-dimensional monitoring data of equipment at the railway catenary site; The collected meteorological data and multi-dimensional monitoring data of the equipment are input into a pre-trained prediction model, which outputs prediction results. Based on the prediction results, the equipment status trend analysis and fault warning are performed.

[0053] Example 4 If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0054] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0055] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0056] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0057] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0058] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for predicting the condition of railway catenary, characterized in that, Includes the following steps: Real-time collection of meteorological data and multi-dimensional monitoring data of equipment at the railway catenary site; The collected meteorological data and multi-dimensional monitoring data of the equipment are input into a pre-trained prediction model, the prediction results are output, and the equipment status trend analysis and fault warning are performed based on the prediction results. The prediction model is built based on historical monitoring data and uses a multi-objective regression analysis model for fault prediction.

2. The method for predicting the condition of a railway catenary according to claim 1, characterized in that, In the step of real-time acquisition of meteorological data and multi-dimensional monitoring data of the railway catenary, the meteorological data includes wind speed, wind direction, temperature, humidity and atmospheric pressure, and the multi-dimensional monitoring data of the equipment includes basic information of the equipment, anchor offset of the catenary, tension, clamp temperature, insulator leakage current, static angle and dynamic amplitude.

3. The method for predicting the condition of a railway catenary according to claim 2, characterized in that, The method for constructing and training the prediction model in the steps of inputting the collected meteorological data and multi-dimensional monitoring data of the equipment into a pre-trained prediction model, outputting prediction results, and performing equipment status trend analysis and fault early warning based on the prediction results is as follows: A historical dataset is acquired, and the data in the historical dataset is cleaned by removing missing values ​​to obtain a preprocessed historical dataset; the historical data includes historical meteorological data and multi-dimensional monitoring data from equipment. The preprocessed historical dataset is divided into feature variables and target variables; The feature variables and target variables are divided into training sets and test sets, a multi-objective regression analysis model is initialized, and the initialized multi-objective regression analysis model is trained using the training set to generate an initial prediction model; The initial prediction model is used to make predictions using the test set to obtain prediction results. The initial prediction model is then trained based on the residuals between the prediction results and the true values ​​to obtain a trained prediction model.

4. The method for predicting the condition of a railway catenary according to claim 3, characterized in that, The multi-objective regression analysis model adopts a multi-output linear regression model, a multi-task learning model, a decision tree-based multi-output regression model, or a deep neural network model.

5. The method for predicting the condition of a railway catenary according to claim 3, characterized in that, The method for dividing the preprocessed historical dataset into feature variables and target variables is as follows: Based on the preprocessed historical dataset, target variables corresponding to the prediction target are selected, and a distribution histogram of the target variables is plotted to analyze their overall distribution characteristics. The other columns of data in the preprocessed dataset, excluding the target variable, are defined as the feature variable set, and the target variable is defined as the prediction label.

6. The method for predicting the condition of a railway catenary according to claim 3, characterized in that, The method for training the initial prediction model in the step of using a test set to predict the initial prediction model, obtaining the prediction result, and training the initial prediction model based on the residual between the prediction result and the true value to obtain a trained prediction model is as follows: Calculate the mean squared error and coefficient of determination between the predicted results and the actual values ​​to quantitatively evaluate the prediction accuracy and interpretability of the initial prediction model under the default hyperparameters. The residual between the predicted result and the actual value is calculated to obtain the residual calculation result. Based on the residual calculation result, a residual distribution histogram is plotted to test whether the error meets the basic assumptions of linear regression, thereby diagnosing whether there is a systematic bias in the initial prediction model. If the initial prediction model is found to have systematic biases, a grid search method is used to traverse the preset hyperparameter space, with the goal of minimizing the mean squared error. The optimal combination of hyperparameters is selected through cross-validation, thereby obtaining a well-trained prediction model.

7. The method for predicting the condition of a railway catenary according to claim 3, characterized in that, The method for analyzing equipment status trends and providing fault warnings based on prediction results is as follows: the prediction results are compared with preset alarm threshold parameters. If the predicted value or the real-time value of the equipment's multi-dimensional monitoring data exceeds the preset alarm threshold parameters, an alarm is triggered and alarm information is generated.

8. A railway catenary condition prediction system, based on the railway catenary condition prediction method according to claim 1, characterized in that, include: The data acquisition module is used to collect meteorological data and multi-dimensional monitoring data of the railway catenary in real time. The prediction module is used to input the collected meteorological data and multi-dimensional monitoring data of the equipment into the pre-trained prediction model, output the prediction results, and perform equipment status trend analysis and fault early warning based on the prediction results. The prediction model is built based on historical monitoring data and uses a multi-objective regression analysis model for fault prediction.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the railway catenary condition prediction method according to any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the railway catenary condition prediction method according to any one of claims 1 to 7.

Images