Fault determination method, train operation scheduling method, vehicle inspection and maintenance scheduling method, and train

Abstract

The present disclosure provides a fault determination method and a train operation scheduling method based on a pantograph-catenary-rail-tunnel system. The fault determination method comprises: on the basis of historical operating environment information, constructing a predetermined graph to represent associations among historical operating environment data, and training a target graph neural network model; on the basis of the predetermined graph, determining target associations between real-time operating state information and real-time operating environment information monitoring data, and inputting the target association into the model; and outputting potential fault information associated with an operating environment, and performing a fault troubleshooting operation. The train operation scheduling method comprises: on the basis of operating state information, operational mileage information, and maintenance information of a target train in a historical time period, generating health state information of the target train; and in view of potential fault information of a reference train, generating an operating policy of the target train, so as to control whether the target train continues to operate on an initial line or is scheduled to operate on an alternative line. In addition, the present disclosure further provides a train operation scheduling method and a vehicle inspection and maintenance scheduling method based on a vehicle-ground integrated system. The train operation scheduling method comprises: on the basis of state monitoring information, operational mileage information, maintenance record information, and historical operating environment information of each component of a plurality of candidate trains, generating health state information of each candidate train; and in view of state information generated by monitoring information of each line to be operated upon, generating an operation scheduling plan of each candidate train, so as to schedule each candidate train to operate on each line to be operated upon. The vehicle inspection and maintenance scheduling method comprises: analyzing historical operational mileage and initial fault diagnosis information of a plurality of vehicles to determine information of each component to be inspected and maintained; on the basis of the operating position and operating speed of each vehicle, separately calculating a traveling duration required for each vehicle to reach a ground inspection and maintenance site; on the basis of the information of each component to be inspected and maintained, each traveling duration, and ground inspection and maintenance operating state information, generating inspection and maintenance scheduling policy information of each vehicle to indicate information of a track on which each vehicle will be located when waiting for inspection and maintenance at the ground inspection and maintenance site; and scheduling each vehicle to travel into the ground inspection and maintenance site according to corresponding track information thereof and wait for inspection and maintenance.

Description

Fault determination methods, train operation scheduling methods, vehicle maintenance scheduling methods, and train Technical Field

[0001] This disclosure relates to the field of rail vehicle operation and management technology, and more specifically, to a fault determination method, a train operation scheduling method, and a train based on a network track tunnel system. Background Technology

[0002] In the traditional electromechanical system integration model, trains, track maintenance, etc. are independent and lack overall planning and management, resulting in data silos between trains, pantographs, catenary, tracks, and tunnels, which leads to insufficient collaborative fault handling capabilities and train operation scheduling capabilities. Summary of the Invention

[0003] One aspect of this disclosure provides a fault determination method based on a track-tunnel network system, comprising: acquiring first real-time operating status information of a target train during a target time period and first real-time operating environment information of the line it travels on, wherein the first real-time operating environment information includes real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data collected by train monitoring equipment at the same time and space as the real-time operating status information; determining target correlation relationships between the real-time operating status information, real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data based on a predetermined graph, wherein the predetermined graph is constructed based on historical pantograph-catenary data, historical track data, and historical tunnel data, and is used to characterize the correlation relationships between the historical pantograph-catenary data, historical track data, and historical tunnel data; inputting the target correlation relationships into a trained target graph neural network model, and outputting potential fault information related to the operating environment of the target train, wherein the potential fault information includes at least potential fault points and fault occurrence probabilities, so as to perform fault elimination operations on potential fault points before the target train travels to the potential fault points.

[0004] Another aspect of this disclosure provides a train operation scheduling method based on a network track-tunnel system, comprising: acquiring potential fault information detected when a reference train is running on an initial line, wherein the potential fault information includes one or more of the following: catenary fault information, track fault information, and tunnel fault information related to the operating environment of the initial line; acquiring first operating status information, first running mileage information, first maintenance information, and multiple first waiting-to-run line information of a target train within a first historical time period, wherein the target train is a train planned to run on the initial line within a target preset time period; generating first health status information of the target train based on the first operating status information, first running mileage information, and first maintenance information; and generating a first operating strategy for the target train based on the first health status information and potential fault information, so as to control the target train to continue running on the initial line or to schedule the target train to run on a first alternative line other than the initial line from the multiple first waiting-to-run line information.

[0005] Another aspect of this disclosure provides a train operation scheduling method based on a vehicle-ground integrated system, comprising: acquiring status monitoring information of each component configured on multiple candidate trains, mileage information of each component, maintenance record information of each component, historical operating environment information of multiple candidate trains, and monitoring information of multiple lines to be operated; generating health status information of each candidate train based on the status monitoring information of each component, mileage information of each component, maintenance record information of each component, and historical operating environment information of each candidate train; generating status information of each line to be operated based on the monitoring information of each line to be operated; and generating an operation scheduling plan for each candidate train based on the health status information of each candidate train and the status information of each line to be operated, so as to schedule each candidate train to run on each line to be operated.

[0006] Another aspect of this disclosure provides a vehicle maintenance scheduling method based on a vehicle-to-ground integrated system, comprising: acquiring the operating status information, initial fault diagnosis information, and ground maintenance operation status information of multiple vehicles; analyzing the historical mileage and initial fault diagnosis information of each vehicle to determine the information of each component to be maintained; calculating the travel time required for each vehicle to reach the ground maintenance site based on the operating location and speed of each vehicle; generating maintenance scheduling strategy information for multiple vehicles based on the information of each component to be maintained, the travel time, and the ground maintenance operation status information; and scheduling each vehicle to enter the ground maintenance site according to its corresponding track information to wait for maintenance based on the maintenance scheduling strategy information. Attached Figure Description

[0007] Figure 1 schematically illustrates an exemplary system architecture to which a fault determination method based on a track-tunnel system can be applied according to an embodiment of the present disclosure;

[0008] Figure 2 schematically illustrates a flowchart of a fault determination method based on a track-tunnel system according to an embodiment of the present disclosure;

[0009] Figure 3 illustrates a method for fault determination based on multiple vehicles according to an embodiment of the present disclosure;

[0010] Figure 4 schematically illustrates a structural block diagram of a fault determination device based on a track-tunnel system according to an embodiment of the present disclosure;

[0011] Figure 5 schematically illustrates an exemplary system architecture to which train operation scheduling methods can be applied according to embodiments of the present disclosure;

[0012] Figure 6 schematically illustrates a flowchart of a train operation scheduling method according to an embodiment of the present disclosure;

[0013] Figure 7 schematically illustrates the working principle of a positioning synchronization system according to an embodiment of the present disclosure;

[0014] Figure 8 schematically illustrates a structural block diagram of a train operation scheduling device according to an embodiment of this application;

[0015] Figure 9 schematically illustrates an exemplary system architecture for applying a train operation scheduling method based on a vehicle-ground integrated system according to embodiments of the present disclosure;

[0016] Figure 10 schematically illustrates the working principle of the novel daily train inspection system according to an embodiment of the present disclosure;

[0017] Figure 11 schematically illustrates a flowchart of a train operation scheduling method based on a vehicle-ground integrated system according to an embodiment of the present disclosure;

[0018] Figure 12 schematically illustrates the process of generating health status information according to an embodiment of the present disclosure;

[0019] Figure 13 schematically illustrates a flowchart of a train operation scheduling method based on a vehicle-ground integrated system according to an embodiment of the present disclosure;

[0020] Figure 14 schematically illustrates a structural block diagram of a train operation scheduling device based on a vehicle-ground integrated system according to an embodiment of the present disclosure;

[0021] Figure 15 schematically illustrates a vehicle maintenance scheduling method applicable to an integrated vehicle system according to an embodiment of the present disclosure;

[0022] Figure 16 schematically illustrates the working principle of the intelligent control platform according to an embodiment of the present disclosure;

[0023] Figure 17 schematically illustrates the working principle of the intelligent rack overhaul system according to an embodiment of the present disclosure;

[0024] Figure 18 schematically illustrates a flowchart of a vehicle maintenance scheduling method based on a vehicle-body integrated system according to an embodiment of the present disclosure;

[0025] Figure 19 schematically illustrates the process of generating a vehicle maintenance scheduling strategy according to an embodiment of the present disclosure;

[0026] Figure 20 schematically illustrates the process of generating vehicle maintenance waiting time according to an embodiment of the present disclosure;

[0027] Figure 21 schematically illustrates an example judgment flowchart for determining the component to be repaired according to an embodiment of the present disclosure;

[0028] Figure 22 schematically illustrates a structural block diagram of a vehicle maintenance and scheduling device based on a vehicle-body integrated system according to an embodiment of the present disclosure;

[0029] Figure 23 schematically illustrates a structural block diagram of a train module according to an embodiment of the present disclosure; and

[0030] Figure 24 schematically illustrates a block diagram of an electronic device suitable for implementing any of the methods described above, according to embodiments of the present disclosure. Detailed Implementation

[0031] The embodiments of this disclosure will be further described below with reference to the accompanying drawings.

[0032] The embodiments of the present disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of the present disclosure.

[0033] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0034] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0035] When using expressions such as "at least one of A, B, and C", they should generally be interpreted in accordance with the meaning that is commonly understood by a person skilled in the art (e.g., "a system having at least one of A, B, and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B, and C, etc.).

[0036] In the embodiments disclosed herein, the collection, updating, analysis, processing, use, transmission, provision, disclosure, and storage of data (e.g., including but not limited to user personal information) comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures have been taken to prevent unauthorized access to user personal information data and to safeguard user personal information security, network security, and national security.

[0037] In the embodiments disclosed herein, user authorization or consent is obtained before acquiring or collecting user personal information.

[0038] In traditional electromechanical system integration models, train operation data, track condition monitoring, and tunnel environmental parameters are scattered, lacking effective data interaction and fusion mechanisms. When a fault occurs, it is difficult to quickly locate the root cause from a global perspective, resulting in low efficiency in collaborative handling.

[0039] Current fault diagnosis relies on human experience or pre-set algorithms, allowing only post-event analysis of fault phenomena. This passive fault handling approach struggles to capture subtle anomalies during system operation, fails to identify early signs of potential faults, and cannot predict the likelihood and scope of a fault's occurrence. In the complex and ever-changing railway operating environment, potential faults can escalate into major safety accidents at any time, yet existing mechanisms are unable to develop targeted prevention and emergency response plans in advance. Once a fault occurs, it can easily trigger a chain reaction of train delays, equipment damage, and other consequences, causing not only huge economic losses to railway transport companies but also seriously threatening passenger safety and the stable operation of railway transportation.

[0040] This disclosure provides a fault determination method based on a track-tunnel network system, comprising: acquiring first real-time operating status information of a target train during a target time period and first real-time operating environment information of the line it travels on, wherein the first real-time operating environment information includes real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data collected by train monitoring equipment at the same time and space as the real-time operating status information; determining target correlation relationships between the real-time operating status information, real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data based on a predetermined graph, wherein the predetermined graph is constructed based on historical pantograph-catenary data, historical track data, and historical tunnel data, and is used to characterize the correlation relationships between the historical pantograph-catenary data, historical track data, and historical tunnel data; inputting the target correlation relationships into a trained target graph neural network model, and outputting potential fault information related to the operating environment of the target train, wherein the potential fault information includes at least potential fault points and fault occurrence probabilities, so as to perform fault elimination operations on potential fault points before the target train travels to the potential fault points.

[0041] Figure 1 schematically illustrates an exemplary system architecture for applying a fault determination method based on a track-tunnel system according to embodiments of the present disclosure. It should be noted that Figure 1 is merely an example of a system architecture to which embodiments of the present disclosure can be applied, to help those skilled in the art understand the technical content of the present disclosure, but does not imply that embodiments of the present disclosure cannot be used in other devices, systems, environments, or scenarios.

[0042] As shown in Figure 1, the system architecture 100A according to this embodiment includes a network track tunnel system 110A, a control center 120A, a train monitoring system 130A, and a positioning and synchronization system 140A.

[0043] The track-rail-tunnel system 110A includes a pantograph-catenary detection system 111A, a track detection system 112A, a tunnel detection system 113A, and a main unit 114A.

[0044] The pantograph-catenary inspection system 111A is configured to acquire catenary parameter data, pantograph status data, and pantograph-catenary relationship data collected by pantograph-catenary inspection equipment installed on the top of the train. Examples include conductor pull-out value detection, pantograph-catenary arc detection, temperature detection, and hard spot detection.

[0045] The track inspection system 112A is configured to acquire track geometric parameter data, track component status data, and train operating environment data collected by track inspection equipment installed on the bottom of the train. The track inspection system 112A also includes: a track condition inspection device, a track geometric parameter detection device, and a rail profile detection device (not shown in the figure). The track condition inspection device can acquire track image information acquired at high speed by the track inspection equipment, and extract damage information such as rail surface scratches, fish scale damage, spalling, wave wear, ballast bed cracks, and missing fasteners from the images. It uses a high-precision camera to dynamically acquire data on the track surface, and then uses digital image processing technology to automatically identify defects in fasteners, rail surfaces, and ballast bed, and transmit faults in real time. The track geometric parameter detection device can establish a navigation coordinate system to achieve relative positioning and attitude measurement of the vehicle. Simultaneously, combined with a vibration compensation device, it achieves vibration compensation and track gauge measurement; combined with a positioning synchronization system, it achieves precise positioning synchronization. Finally, it outputs track geometric parameter information such as track gauge, superelevation, track direction, and elevation in real time. The rail profile detection device can perform real-time, dynamic, and high-precision detection of the entire rail cross-section and wear. It automatically establishes a standard rail profile, matches and analyzes each collected profile with the standard profile, and outputs vertical and horizontal wear results in real time.

[0046] The tunnel inspection device 113A is configured to acquire tunnel clearance data and tunnel structural status data collected by tunnel inspection equipment installed at the front of a train or in a tunnel. Specifically, a high-speed laser scanner can be used to quickly scan the tunnel surface to obtain three-dimensional information of the tunnel cross-section, presenting the spatial morphology of the tunnel. The high-speed acquired tunnel cross-section data is compared with a preset subway clearance model (which specifies the dimensions and other standards that subway tunnel spaces must meet) to determine whether the actual tunnel space meets the standards, obtain the tunnel clearance inspection results, and check for problems such as encroachment. The acquired two-dimensional images are used to identify surface defects in the tunnel, such as cracks and spalling, to assess the tunnel's health status.

[0047] The main unit 114A is usually installed in the car cabinet or under the seat, and is configured to collect, organize, analyze, determine faults and provide early warnings, and store data collected by the pantograph-catenary detection device 111A, track detection device 112A, and tunnel detection device 113A.

[0048] Control center 120A includes a track maintenance management system 121A, a train dispatching system 122A, other related subsystems, and terminal equipment (not shown in the diagram). The track maintenance management system 121A is configured to receive track and tunnel data transmitted from the track and tunnel system 110A, and to classify, store, and analyze this data for later retrieval. The train dispatching system 122A is configured to monitor train positions and status in real time, formulate and adjust train operation strategies, and coordinate train passing and avoidance maneuvers. The terminal equipment provides a visual interface to display data processing and analysis results, early warning information, etc., to track maintenance personnel, enabling them to monitor vehicle and environmental information in real time and develop contingency plans for potential faults.

[0049] The train monitoring system 130A monitors the train's operating status in real time, such as speed and direction of travel; the status of train equipment, such as traction system, braking system, bogies, and doors; and also includes: train number information, train mileage, train fault diagnosis information, train signaling equipment PHM (Prognostics and Health Management) data and related diagnostic information. The train PHM diagnostic data can be used to guide daily train inspections. The positioning and synchronization system 140A obtains train location information through RFID tags installed at specific locations on the track or via satellite positioning.

[0050] Figure 2 schematically illustrates a flowchart of a fault determination method based on a track-tunnel system according to an embodiment of the present disclosure.

[0051] As shown in Figure 2, method 200A includes operations S210A to S230A.

[0052] During operation S210A, the first real-time operating status information of the target train and the first real-time operating environment information of the line it travels on are acquired within the target time period. The first real-time operating environment information includes real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data collected by the train monitoring equipment at the same time and space as the real-time operating status information.

[0053] According to embodiments of this disclosure, the target train can be a currently operating train. The first real-time operating status information reflects the real-time operating status of the target train itself, such as whether the train is running smoothly, whether it is unstable, whether it is vibrating, and the train's speed and acceleration. The first real-time operating environment information reflects the environmental conditions of the track on which the target train operates, including at least real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data. Specifically, the real-time pantograph-catenary monitoring data reflects the working status of the pantograph-catenary system, such as the voltage, current, and contact pressure of the catenary; the real-time track monitoring data includes, for example, the geometric dimensions, wear condition, and vibration of the track; and the real-time tunnel monitoring data includes, for example, the temperature, humidity, and structural deformation within the tunnel.

[0054] During S220A operation, based on a predetermined map, the target correlation relationships between real-time operating status information, real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data are determined. The predetermined map is constructed based on historical pantograph-catenary data, historical track data, and historical tunnel data, and is used to characterize the correlation relationships between these data.

[0055] According to embodiments of this disclosure, the predetermined map is constructed based on historical pantograph-catenary data, historical track data, and historical tunnel data, and may also include historical full-scale train operation status information collected by the train monitoring system. The historical pantograph-catenary data, historical track data, historical tunnel data, and historical full-scale operation status information serve as nodes in the map. The relationships between the nodes are connected by edges. For example, there is an edge representing "power transmission" between the vehicle and the pantograph-catenary system, an edge representing "physical contact" between the vehicle and the track, an edge representing "spatial location" between the vehicle and the tunnel, and an edge representing "indirect influence" between the pantograph-catenary system and the track, etc. Each edge also has attribute information to describe the strength, frequency, confidence level, etc., of the relationship. For example, for the "physical contact" relationship edge between the vehicle and the track, its attributes may include the degree of influence of track conditions on the wear of vehicle components. Through the predetermined map, the interrelationships between real-time data can be uncovered, enabling a more comprehensive understanding of the interactions between various factors during train operation.

[0056] During operation S230A, the target association relationship is input into a trained target graph neural network model, which outputs potential fault information related to the operating environment of the target train. This potential fault information includes at least the potential fault point and the probability of fault occurrence, so that troubleshooting can be performed on the potential fault point before the target train reaches it.

[0057] According to embodiments of this disclosure, the target graph neural network is a neural network used to process data with a graph structure, capable of learning and analyzing relationships between data. The trained model can output potential fault information related to the operating environment of the target train based on the input target associations. The potential fault information includes at least potential fault points, i.e., locations where faults may occur, such as section A of the track, and the corresponding probability of fault occurrence, such as a 76% probability of the train skidding when traveling to section A of the track. Therefore, by obtaining potential fault information in advance, targeted troubleshooting operations can be performed on potential fault points before the target train reaches them, thereby preventing faults and ensuring the safe operation of the train.

[0058] According to embodiments of this disclosure, by collecting real-time operating status information and real-time operating environment information of the target train, combining this information with a predetermined graph to determine the correlation between various data points, and using a trained graph neural network model for analysis, potential fault information, including potential fault points and the probability of fault occurrence, can be output in advance. Therefore, staff can take measures before the train reaches the potential fault point to mitigate the impact of the fault or even prevent its occurrence. The predetermined graph is constructed based on historical data, which can fully utilize the entire historical data to mine potential correlations between different data points. Based on this, the target correlations of real-time data are determined, providing more accurate relationships for fault prediction and improving the accuracy of the fault determination method. Furthermore, on the one hand, timely acquisition of potential fault information and early execution of fault-solving operations can effectively reduce the risk of faults occurring during train operation, ensure the safe operation of the train, reduce train delays and stoppages caused by faults, improve the reliability and stability of railway transportation, and protect the lives and property of passengers. On the other hand, based on potential fault information, staff can rationally allocate resources such as maintenance personnel, equipment, and spare parts, avoiding resource waste and over-configuration. Targeted preparations before a failure occurs can ensure that resources can be put into use in a timely manner when needed, thus improving resource utilization efficiency.

[0059] According to embodiments of this disclosure, in addition to the graph neural network model mentioned herein, support vector machines (SVM), random forests, and other neural network models can also be used, and training can be performed using sample data and labels. Real-time vehicle operation status data and track and tunnel data are input into the trained model. Based on the learned correlations, the real-time data is predicted and analyzed to determine whether each data point belongs to the normal range. For example, for track elevation data, if the model predicts that the data of a certain measurement point deviates significantly from the normal elevation range, it is marked as a potential anomalous data point. Cluster analysis is performed on the identified anomalous data points. Based on the spatial location (e.g., track mileage) and temporal correlation of the data points, adjacent anomalous data points are clustered into potential problem areas. For example, if multiple consecutive measurement points in a certain section of the track are marked as anomalous, and these anomalous data points also have a certain temporal continuity (e.g., appearing within the same time period), then that track section is marked as a potential problem area. Simultaneously, a priority is assigned to each problem area based on the type and severity of the anomalous data for subsequent maintenance and processing. Output the marked potential problem areas and their related information (such as location, anomaly type, priority, etc.).

[0060] According to embodiments of this disclosure, the method for constructing a predetermined map includes steps 11 to 16.

[0061] Step 11: Obtain the historical operating status information and historical operating environment information of the target train within the historical time period. The historical operating status information includes historical driving stability data, and the historical operating environment information includes historical pantograph-catenary data, historical track data, and historical tunnel data collected by the train monitoring equipment at the same time and space as the historical operating status information.

[0062] According to embodiments of this disclosure, the historical time period can be any time period before constructing the preset map. When setting the historical time period, in order to collect more comprehensive data, all factors that affect trains and the environment should be covered as much as possible (such as weather conditions, holiday information, passenger flow, etc.). It is generally believed that the longer the historical time period is set, the more comprehensive the data collection will be.

[0063] According to embodiments of this disclosure, vehicle operation data is acquired from a train monitoring system, including but not limited to real-time vehicle speed, acceleration, stability indicators, instability status, and vibration data; data on pantograph-catenary contact force, current, voltage, and slide plate wear are collected from a pantograph-catenary inspection device; track geometric parameters (e.g., gauge, elevation irregularities, and track alignment deviation), rail profile, and corrugation data are collected from a track inspection device; and data on tunnel temperature, humidity, ventilation status, and lining structure status are acquired using a tunnel inspection device. Simultaneously, the timestamps and spatial location information of each data acquisition are recorded to ensure spatiotemporal consistency of the data.

[0064] According to embodiments of this disclosure, relevant data on vehicle, pantograph-catenary, track, and tunnel malfunctions are collected from historical data, including malfunction type, malfunction time, malfunction location, and operational and environmental data for a period before and after the malfunction. The malfunction data is labeled and distinguished from normal operation data to form a complete dataset.

[0065] According to embodiments of this disclosure, noisy data, outliers, and missing values ​​are removed from the data. Data of different types and dimensions are normalized to eliminate the impact of differences in dimensions between data on subsequent analysis.

[0066] Step 12: Based on the L types of data contained in historical driving stability data, historical pantograph-catenary data, historical track data, and historical tunnel data, construct at least one transaction dataset, wherein the at least one transaction dataset contains L transactions corresponding to the L types of data.

[0067] According to embodiments of this disclosure, historical driving stability data may include M types of data, including at least: smoothness detection data of the target train, instability detection data, vibration detection data, and idling coasting data.

[0068] According to embodiments of this disclosure, historical pantograph-catenary data may include N types of data, including at least: target train catenary pressure data, catenary current data, pantograph-catenary contact pressure data, track arcing rate data, and pantograph-catenary foreign object data.

[0069] According to embodiments of this disclosure, historical track data may include X types of data, including at least: rail profile data, rail corrugation data, turnout abnormal status data, track gauge detection data, and rail surface status data during idle sliding.

[0070] According to embodiments of this disclosure, historical tunnel data may include Y types of data, including at least: tunnel foreign object intrusion data, tunnel clearance data, and tunnel video anomaly data.

[0071] Where M, N, X, and Y are all positive integers greater than or equal to 1, and M+N+X+Y=L, meaning that a transaction dataset includes at least 4 transactions corresponding to 4 data types.

[0072] According to embodiments of this disclosure, at least one transaction dataset further includes L state categories corresponding to transactions. The state category corresponding to a transaction is used to characterize the degree represented by the data value of that type of data. For example, for vehicle speed data, the state categories can be divided into several intervals such as low speed, medium speed, and high speed; for track irregularity data, the state categories are divided into categories such as slight, moderate, and severe based on their severity.

[0073] For ease of understanding, the status categories of the above data are exemplarily divided into three levels.

[0074] According to the embodiments of this disclosure, the stability test data (excellent / good / poor) is as follows: excellent indicates that all stability indicators during train operation are within the ideal range, with no obvious shaking or bumping; good indicates that there are slight signs of instability, but do not exceed the specified threshold; poor indicates that the stability indicators are obviously abnormal, and the train shakes and bumps violently.

[0075] According to embodiments of this disclosure, instability detection data (low / medium / high) are provided, where low indicates that the detected instability signal is weak and the possibility of train instability is extremely low; medium indicates that there is a certain degree of instability signal and the train may experience slight instability; and high indicates that the instability signal is strong.

[0076] According to embodiments of this disclosure, vibration detection data (weak / medium / strong) are as follows: weak indicates that the train vibration amplitude is small and within the normal vibration range, which will not have an adverse effect on train equipment and passengers; medium indicates that the vibration amplitude has increased, exceeding the normal range but not reaching a serious level, which may have a certain impact on some precision equipment; strong indicates that the train vibration is severe, which may lead to equipment loosening, damage, or even affect the train's operational safety and stability.

[0077] According to embodiments of this disclosure, the idling and coasting data (weak / medium / strong) are as follows: weak indicates that the degree of train idling and coasting is slight and has little impact on the train's operating speed and power output; medium indicates that the idling and coasting is more obvious and has a certain impact on the train's operation; strong indicates that the idling and coasting is severe, the train's power drops significantly, the operating speed decreases significantly, and it may even cause the train to stop.

[0078] According to embodiments of this disclosure, the contact network voltage data (low / medium / high) indicates that the contact network voltage is below the lower limit of the normal operating range, which may lead to insufficient train power; medium indicates that the contact network voltage is within the normal fluctuation range; and high indicates that the contact network voltage is above the upper limit of the normal operating range, which may damage the train's electrical equipment.

[0079] According to embodiments of this disclosure, the network current data (low / medium / high) indicates that the current intensity is low, which means that the train is under light load or there is an electrical system abnormality, resulting in insufficient power output; medium indicates that the network current is at a normal level, which meets the power consumption requirements of the train under the current operating conditions; high indicates that the current is too high, which may be caused by excessive train load, electrical equipment failure or contact network abnormality.

[0080] According to embodiments of this disclosure, the pantograph-catenary contact pressure data (low / medium / high) indicates that the contact pressure is too low, which may lead to poor contact between the pantograph and the catenary, resulting in phenomena such as disconnection and arcing; medium indicates that the contact pressure is moderate and can ensure good contact between the pantograph and the catenary; high indicates that the contact pressure is too high, which will increase the wear of the pantograph and the catenary, shorten the service life of the equipment, and may even damage the pantograph and the catenary.

[0081] According to embodiments of this disclosure, the arcing rate data of the line is (low / medium / high). Low indicates that arcing occurs less frequently, the pantograph-catenary system is in good operating condition, and the wear and tear on the equipment is small. Medium indicates that there is a certain frequency of arcing, and its development trend needs to be monitored and timely maintenance and adjustment should be carried out. High indicates that arcing occurs frequently, which will accelerate the wear of the pantograph and the contact network, reduce equipment reliability, and may cause power supply failure.

[0082] According to embodiments of this disclosure, foreign object data for the pantograph-catenary system is categorized as low / medium / high. Low indicates that the detected foreign objects have little impact on the pantograph-catenary system, or there are almost no foreign objects. Medium indicates that there are a certain number or volume of foreign objects, which may interfere with normal pantograph-catenary contact. High indicates that there are many foreign objects or that are large in size, which seriously threaten the safety of the pantograph-catenary system and may cause damage to the pantograph or contact wire failure.

[0083] According to embodiments of this disclosure, rail profile data (excellent / good / poor) are categorized as follows: "excellent" indicates that the profile is very close to the standard profile and the rail wear is minimal; "good" indicates that there is some wear or deformation, but it does not affect the normal operation of the train; and "poor" indicates that the profile changes significantly, which may affect the stability and safety of the train.

[0084] According to embodiments of this disclosure, rail corrugation data (heavy / medium / light) are provided, where "light" indicates a light degree of corrugation with minimal impact on train operation; "medium" indicates a moderate degree of corrugation that may cause some vibration in the train; and "heavy" indicates severe corrugation that significantly affects the smoothness of train operation and may even exacerbate damage to wheels and rails.

[0085] According to embodiments of this disclosure, the abnormal status data of the turnout (none / minor / major) is as follows: "none" indicates that the turnout is in normal condition and can perform functions such as switching and locking normally; "minor" indicates that the turnout has some minor problems, such as slight wear of components or indication errors, but does not affect the basic function of the turnout or the passage of trains; "major" indicates that the turnout has serious problems, such as the switch rail not fitting tightly or the switch machine malfunctioning, which may cause the turnout to be unusable and affect train operation.

[0086] According to embodiments of this disclosure, track gauge detection data (standard / near standard / deviation from standard) are provided, where "standard" means that the track gauge fully complies with the prescribed standard value and can provide a good running track for the train; "near standard" means that the track gauge deviates from the standard value to a certain extent, but within the allowable range and has little impact on train operation; "deviation from standard" means that the track gauge deviation exceeds the allowable range, which may lead to unstable train operation or even the risk of derailment.

[0087] According to embodiments of this disclosure, the rail surface condition data (dry / wet / oiled) during idling and coasting includes a "dry" rail surface that provides better adhesion conditions, which helps the train return to normal operation; a "wet" rail surface that reduces the adhesion coefficient and increases the possibility of idling and coasting; and an "oiled" rail surface that severely reduces the adhesion coefficient, exacerbating the idling and coasting situation and making it difficult to return to normal operation.

[0088] According to embodiments of this disclosure, tunnel foreign object intrusion data (none / small amount / large amount) are provided, where "none" indicates that no foreign objects have intruded into the clearance in the tunnel; "small amount" indicates that some small foreign objects have intruded into the clearance, but do not pose a serious threat to train operation safety; and "large amount" indicates that a large number or large foreign objects have intruded into the clearance, which may collide with the train and seriously threaten train operation safety.

[0089] According to embodiments of this disclosure, tunnel clearance data (compliant / near critical / exceeding) means that the tunnel clearance fully meets the requirements for safe train operation; "near critical" means that the tunnel clearance is close to the minimum allowable value, which may affect the operation of some oversized freight trains or trains under special circumstances; "exceeding" means that the tunnel clearance exceeds the specified range, which may lead to dangerous situations such as the train scraping against the tunnel wall.

[0090] According to the embodiments of this disclosure, the tunnel video abnormal data (none / minor abnormality / serious abnormality) is categorized as follows: "none" indicates that the tunnel video monitoring screen is normal and no abnormalities are found; "minor abnormality" indicates that there are some minor abnormalities in the video that do not affect the safety of train operation, such as flickering lights or partial blurring of the image; "serious abnormality" indicates that there are obvious abnormalities in the video, such as fire or collapse, which would pose a serious threat to the safety of train operation.

[0091] According to embodiments of this disclosure, one or more time windows are obtained from a historical time period. The preprocessed data is divided into multiple transactions according to the time windows. For example, if the historical time period is two years, a year, a month, a day, or an hour can be used as the time window. Each transaction includes transactions of all data types related to vehicles, pantographs, tracks, and tunnels within the same time window, as well as the corresponding status category of that transaction.

[0092] For example, within a time window of 12:00-17:00, multiple transaction datasets are constructed based on the acquired historical full data.

[0093] Transaction dataset 1: Vehicle A traveling on track A [stability (good), net flow data (low), pantograph-catenary contact pressure data (low), rail profile data (good)].

[0094] Transaction dataset 2: Vehicle B traveling on line A [stability (good), network flow data (low), pantograph-catenary contact pressure data (low), rail profile data (good), turnout anomaly (minor anomaly) and tunnel video anomaly data (none)].

[0095] Transaction dataset 3: Vehicle C traveling on track A [stability (good), network flow data (low), pantograph-catenary contact pressure data (low), rail profile data (good), turnout anomaly (minor anomaly) and tunnel video anomaly data (none), rail surface condition data during idle sliding (dry)].

[0096] Step 13: Determine multiple frequent itemsets based on the support and support threshold of each transaction.

[0097] According to embodiments of this disclosure, the support of each transaction in each transaction dataset is calculated and compared with a support threshold. If the support threshold is set as follows: the frequency of occurrence in the above transaction data reaches 50% of the total number of transaction datasets (3 transaction datasets) (i.e., the frequency is greater than or equal to 2 times) as the criterion for determining a frequent itemset (if it is in the form of a rank, it is assumed that the frequency rank threshold for each transaction is "at least 2 times"). The transactions "Stability (Good)", "Network Flow Data (Low)", and "Rail Profile Data (Good)" all occur 3 times, which is greater than or equal to 50% of the total number of transaction datasets (i.e., 2 times), so the itemsets containing these transactions are frequent itemsets.

[0098] Step 14: Generate multiple frequent sub-itemsets corresponding to each frequent item set based on each transaction in each frequent item set.

[0099] According to embodiments of this disclosure, for a frequent itemset, all its non-empty subsets are frequent itemsets.

[0100] The transaction dataset 1: [Stability (Good), Network Flow Data (Low), Pantograph-Catenary Contact Pressure Data (Low), Rail Profile Data (Good)] is simplified.

[0101] Simplified to: [A (Good), B (Low), C (Low), D (Good)].

[0102] The frequent itemsets of transaction dataset 1 include: [A(Good)], [B(Low)], [C(Low)], [D(Good)], [A(Good), B(Low)], [B(Low), C(Low)], [A(Good), B(Low), C(Low), D(Good)], etc.

[0103] Step 15: Calculate the lift of each frequent subset and determine the relationships between transactions in the frequent subset based on the lift.

[0104] According to an embodiment of this disclosure, exemplarily, in addition to transaction dataset 1, there are also transaction dataset 4: [A(Excellent), B(High), C(Low), D(Poor)] and transaction dataset 5: [A(Good), B(Low), C(High), D(Medium)]. The lift of frequent itemsets B(Low) and C(Low) is calculated respectively: In the three transaction datasets, B(Low) appears twice, so P(B(Low)) = 2 / 3; In the three transaction datasets, C(Low) appears twice, so P(C(Low)) = 2 / 3; The simultaneous occurrence of B(Low) and C(Low) occurs twice in transaction datasets 1 and 5, so P(B(Low)∪C(Low)) = 2 / 3. The lift of B(Low) to C(Low) is given by Equation (1): Lift(I, J) = P(I∪J) / (P(I)×P(J))……Equation (1)

[0105] Where I represents a frequent subset, J represents another frequent subset, and Lift(I, J) represents the lift of the frequent subset I to the frequent subset J.

[0106] According to equation (1), Lift(B(low), C(low)) = 1.5. When the lift is greater than 1, it indicates a positive correlation between B(low) and C(low), meaning that when B is in a "low" state, C is more likely to be in a "low" state. In this example, it shows a positive correlation between grid current data (low) and pantograph-catenary contact pressure data (low), meaning that when grid current is in a "low" state, pantograph-catenary contact pressure is more likely to be in a "low" state. This correlation may reflect a problem in the power supply system that causes both to be low at the same time, and can be used as a reference for detecting anomalies or faults.

[0107] According to embodiments of this disclosure, for each frequent item set, all possible association rules are generated. By calculating and analyzing the lift of different frequent item sets, the association relationships between various transactions in the full historical network track and tunnel data are calculated. The generated association rules are further filtered to remove redundant and meaningless rules.

[0108] Step 16: Construct a predetermined graph for each transaction based on the relationships between them.

[0109] According to embodiments of this disclosure, vehicles, pantograph-catenary systems, tracks, and tunnels are used as nodes in a knowledge graph. Each node contains attribute information. For example, a vehicle node includes attributes such as vehicle ID, vehicle type, service life, real-time operating status (speed, acceleration, stability, etc.), and historical fault records; a pantograph-catenary system node includes attributes such as pantograph-catenary system ID, model, material, real-time contact force, current, voltage, and historical maintenance records; a track node includes attributes such as track ID, laying location, laying material, service life, real-time geometric parameters (gauge, elevation irregularities, etc.), and historical maintenance records; and a tunnel node includes attributes such as tunnel ID, structure type, construction time, real-time environmental parameters (temperature, humidity, ventilation, etc.), and past accident records. Preprocessed data and the associations mined using the aforementioned algorithm are added as node attributes to the corresponding nodes. For example, a vehicle-related association rule (such as "if the vehicle speed is high and the track gauge deviates significantly, then the vehicle vibration intensifies") is used as an attribute of the vehicle node, recording the specific content and confidence level of the association rule.

[0110] According to embodiments of this disclosure, edges between nodes are determined based on the association rules mined by the above algorithm. Edges represent the association relationships between nodes, such as an edge representing "power transmission" between the vehicle and the pantograph-catenary system, an edge representing "physical contact" between the vehicle and the track, an edge representing "spatial location" between the vehicle and the tunnel, an edge representing "indirect influence" between the pantograph-catenary system and the track (influenced through vehicle operation), an edge representing "environmental impact" between the pantograph-catenary system and the tunnel, and an edge representing "structural interaction" between the track and the tunnel. Attributes are added to each edge to describe the strength, frequency, confidence level, and other information of the relationship. For example, for the "physical contact" relationship edge between the vehicle and the track, its attributes may include the degree of influence of track conditions on the wear of vehicle components (obtained based on historical data statistics), the confidence level of this relationship in the algorithm's association rules, etc.

[0111] According to embodiments of this disclosure, a professional graph database is selected to store the knowledge graph. Simultaneously, a real-time update mechanism for the knowledge graph is established, promptly updating the attributes of corresponding nodes and the relationships between edges when new operational or faulty data is generated. Furthermore, the aforementioned association mining algorithm is periodically re-run to update the association rules, thereby optimizing and expanding the knowledge graph.

[0112] According to embodiments of this disclosure, by calculating support, the frequency of occurrence of each data item (vehicle, network, track, and tunnel) in historical data can be determined. Based on this, calculating lift can reveal the correlation strength between different data items, thereby accurately identifying which combinations of vehicle, network, track, and tunnel data are closely related to anomalies. For example, when network flow data is within a specific range, and rail profile data exhibits a certain characteristic, it may show a strong correlation with anomaly data in tunnel video, facilitating the rapid identification of key factor combinations that may trigger anomalies.

[0113] According to embodiments of this disclosure, a knowledge graph constructed based on the mined relationships can be used to build predictive models. When a specific state is detected in a data item, the relationships can be used to predict potential anomalies or faults in the vehicle, network, track, and tunnel systems, thereby issuing timely warnings. For example, if fluctuations in network voltage data are detected and match a combination of network flow data and vehicle stability data that previously caused abnormal track surface wear in historical data, measures can be taken in advance, such as strengthening track surface inspection and maintenance, to prevent faults from occurring or mitigate their impact.

[0114] According to embodiments of this disclosure, the target graph neural network model training steps include steps 21 and 22.

[0115] Step 21: Obtain the predetermined map and historical fault diagnosis information within the historical time period.

[0116] Step 22: Use the predetermined graph as historical sample data and the historical fault diagnosis information as labels to optimize the initial graph neural network model and obtain the target graph neural network model.

[0117] According to embodiments of this disclosure, a predetermined graph is obtained, along with historical fault diagnosis information recording various faults occurring in the vehicle, network, track, and tunnel systems over a historical time period. This information includes the fault type (e.g., minor, moderate, severe), the location of the fault (potential fault point), and the probability of the fault occurring. The obtained graph is used as historical sample data, and the historical fault diagnosis information is used as labels. These are input into an initial graph neural network model. The model makes predictions based on the input graph and then compares the prediction results with the labels (historical fault diagnosis information) to calculate the prediction error. The error is propagated from the output layer to the input layer using a backpropagation algorithm, adjusting the model parameters (e.g., node weights, edge weights) to gradually approximate the true labels. After multiple iterations of training, the model parameters are continuously optimized until the prediction results meet the training termination condition, resulting in the target graph neural network model.

[0118] According to embodiments of this disclosure, in complex environments like rail-tunnel systems where multiple factors interact, traditional fault diagnosis methods often rely on manual experience or simple rule-based judgments, which are not only inefficient but also prone to misjudgments. However, by using a predetermined graph as historical sample data and combining it with historical fault diagnosis information as labels to train an initial graph neural network model, the model can effectively learn the intrinsic connections between different data relationships and faults. When faced with new operational data, the model can more accurately predict potential faults, take preventative measures in advance, and ensure the safe and stable operation of the system. Simultaneously, the target graph neural network model, trained and optimized with a large amount of historical data, can quickly analyze the input relationship graph and derive fault diagnosis results, significantly shortening fault diagnosis time, improving diagnostic efficiency, helping to quickly locate the cause of faults, and reducing adverse impacts on operations. Furthermore, the predetermined graph comprehensively describes the complex relationships between various data in the system, enabling the target graph neural network model to better adapt to the characteristics of complex systems, process multi-source heterogeneous data, deeply mine the potential relationships between data, accurately capture abnormal conditions in system operation, and significantly improve the ability to diagnose and predict faults in complex systems.

[0119] Figure 3 shows a schematic diagram of a method for fault determination based on multiple vehicles according to an embodiment of the present disclosure.

[0120] According to embodiments of this disclosure, the fault determination method based on the network track tunnel system further includes: determining a target fault point from potential fault points based on second real-time operating status information of multiple reference trains in the same area, second real-time operating environment information, and fault occurrence probability, so as to eliminate the fault at the target fault point in advance.

[0121] According to an embodiment of this disclosure, as shown in Figure 3, track and tunnel data of trains 1, 2, and 3 are simultaneously acquired and sent to the track maintenance management system 121A. The system comprehensively considers data from multiple reference trains and verifies and supplements fault information based on the operating conditions of different trains. For example, if multiple trains exhibit similar abnormal operating conditions (such as increased vibration) while passing through a certain area, and the track and tunnel environment information for that area also shows some abnormal indicators (such as increased rail profile deviation), and the probability of a potential fault point in that area being faulty is high, then this potential fault point is more likely to become the target fault point. The train dispatching system 122A generates an operating strategy and dispatches each vehicle based on the potential fault information and the track and tunnel data of multiple trains.

[0122] According to embodiments of this disclosure, the stability detection data of the target vehicle, combined with the detection data of turnouts and track irregularities of a reference train, identifies abnormal track conditions such as turnouts in the same section, and performs fusion diagnosis, which can improve the accuracy of locating abnormal sections. Simultaneously, by combining train cluster analysis, it can identify whether the cause is track-related or vehicle-related, enabling precise maintenance.

[0123] According to embodiments of this disclosure, by combining diagnostic results data such as instability alarms or instability harmonic characteristics of multiple trains and multiple cars in the same section, and data such as rail profile, rail wear, and track gauge detection from the comprehensive inspection vehicle, multi-source data can be used for fusion diagnosis. This can improve the accuracy of identifying abnormal vehicle and track conditions for lateral instability phenomena.

[0124] According to embodiments of this disclosure, vehicle grid voltage and current data are combined with test data such as the static current-carrying capacity of the testing vehicle, pantograph-catenary contact pressure, and arcing rate of the line for integrated diagnosis. By using multiple testing methods of the testing vehicle, supplemented by real-time grid voltage and current data during vehicle operation, the pantograph-catenary current-collecting performance is comprehensively judged, and the pantograph-catenary matching relationship is studied.

[0125] According to the embodiments of this disclosure, when a vehicle detects a foreign object encroaching on the tunnel, affecting the safety of train operation, it immediately feeds back the tunnel encroachment data (location, video images before and after the encroachment location) to the control center 120A for confirmation, and takes timely fault handling measures to ensure the safety of subsequent train operation.

[0126] According to the embodiments of this disclosure, based on the real-time sharing of multi-vehicle track and tunnel detection data, the tunnel clearance data detected by multiple vehicles at the same location are fused and analyzed to determine whether there are any foreign objects gradually approaching the vehicle operation clearance. If so, the tunnel encroachment data (location, video images before and after the encroachment location) is fed back to the control center for confirmation, and fault handling measures are taken in a timely manner to ensure the safety of subsequent train operation.

[0127] According to the embodiments of this disclosure, when a foreign object is detected in the pantograph and catenary, affecting train operation, the foreign object data (location, video images before and after the foreign object location) is immediately fed back to the control center 120A for confirmation. Timely fault handling measures are taken to ensure the safety of subsequent train operation. By optimizing the train control strategy, the pantograph is lowered in the foreign object area and then raised again after sliding over the foreign object location to ensure operational efficiency under abnormal conditions.

[0128] According to embodiments of this disclosure, when a vehicle experiences wheel spin or coasting, track detection data analysis is triggered to determine if it is due to slippery track surface. The track detection data for this location is then uploaded to the control center 120A for subsequent fault investigation. If it is determined that the slippery track surface is the cause, the data is promptly shared with other vehicles. When other vehicles reach this location, they can reduce traction or braking force to minimize wheel spin and coasting. Simultaneously, the relationship between vehicle braking slippage data and track surface conditions is determined, such as the impact of water or oil on the vehicle's braking distance.

[0129] According to embodiments of this disclosure, once the target fault point is identified, troubleshooting can be performed in advance. This enables the implementation of appropriate measures to eliminate potential fault hazards before the fault actually occurs, preventing the fault from affecting train operation and ensuring the normal operation of the rail-tunnel system and the safety of trains.

[0130] According to embodiments of this disclosure, the potential fault information further includes the potential fault type and the expected occurrence time of the potential fault, and the method further includes steps 31 to 32.

[0131] Step 31: Determine the severity and scope of impact of the potential fault based on the potential fault location, potential fault type, and expected occurrence time of the potential fault.

[0132] Step 32: Determine the handling priority and handling strategy for potential faults based on their severity and scope of impact.

[0133] According to embodiments of this disclosure, regarding potential fault points, if the potential fault point is located in a critical part of the rail-tunnel system, such as the track near transportation hubs or the overhead contact line of major power supply lines, a fault in this area may have a significant impact on the operation of the entire system, and the severity will be correspondingly higher. For example, a potential fault at a switch on a busy railway trunk line may cause delays or cancellations of multiple trains. Regarding fault types, different fault types have different degrees of severity. For example, a fault like a broken overhead contact line is much more serious than a minor contact failure, as it may directly prevent trains from obtaining power, causing widespread train stoppages. Severe track deformation faults pose a greater threat to train operation safety than general track wear faults. Regarding the expected occurrence time of a potential fault, if the potential fault is expected to occur during peak train operation periods, its impact may be wider because more trains are involved, causing greater disruption to transportation order and thus increasing its severity. Conversely, if it is expected to occur during off-peak train operation periods, the impact will be relatively smaller, and the severity may be lower.

[0134] According to embodiments of this disclosure, high-severity potential faults are given a higher priority for handling because they pose a greater threat to the rail-tunnel system and train operation safety. These faults should be addressed as quickly as possible to avoid serious consequences. For example, potential faults involving severe track deformation that could lead to train derailment must be prioritized. Potential faults with a wide impact also require priority handling to minimize disruption to the entire transportation system. For instance, overhead contact line faults that could affect train operation on multiple lines should be prioritized for handling by maintenance personnel and resources.

[0135] Based on the determined processing priorities, corresponding processing strategies should be formulated. For high-priority potential faults, a professional maintenance team should be immediately organized for emergency repairs, and sufficient maintenance equipment and spare parts should be allocated. For low-priority potential faults, processing can be scheduled for a suitable time (such as during train downtime), or temporary monitoring measures can be taken to observe their development before deciding on the subsequent processing method. For example, for some minor track wear potential faults, monitoring can be strengthened first, and the wear condition can be checked regularly. Processing can only be carried out when the wear reaches a certain level.

[0136] According to the embodiments of this disclosure, after the fault is repaired, a second test is performed to confirm that the fault has been completely eliminated. The fault handling process, including the cause of the fault, the handling method, and the scheduling effect, is comprehensively summarized and stored in the case library.

[0137] Figure 4 schematically illustrates a structural block diagram of a fault determination device based on a track-tunnel system according to an embodiment of the present disclosure.

[0138] As shown in Figure 4, the fault determination device 400A includes a first acquisition module 410A, a first determination module 420A, and an output module 430A.

[0139] The first acquisition module 410A is configured to acquire the first real-time operating status information of the target train during a target time period and the first real-time operating environment information of the line it travels on. The first real-time operating environment information includes real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data collected by the train monitoring equipment at the same time and space as the real-time operating status information.

[0140] The first determining module 420A is configured to determine the target correlation between real-time operating status information, real-time pantograph-catenary monitoring data, real-time track monitoring data, and real-time tunnel monitoring data based on a predetermined map. The predetermined map is constructed based on historical pantograph-catenary data, historical track data, and historical tunnel data, and is used to characterize the correlation between these data.

[0141] The output module 430A is configured to input the target association relationship into a trained target graph neural network model and output potential fault information related to the operating environment of the target train. This potential fault information includes at least the potential fault point and the probability of fault occurrence, so that troubleshooting can be performed on the potential fault point before the target train reaches it.

[0142] According to embodiments of this disclosure, the method for constructing a predetermined map includes a second acquisition module, a first construction module, a second determination module, a generation module, a calculation module, and a second construction module.

[0143] The second acquisition module is configured to acquire historical operating status information and historical operating environment information of the target train within a historical time period. The historical operating status information includes historical driving stability data, and the real-time operating environment information includes historical pantograph-catenary data, historical track data, and historical tunnel data collected by the train monitoring equipment at the same time and space as the historical operating status information.

[0144] The first construction module is configured to construct at least one transaction dataset based on L data types contained in historical driving stability data, historical pantograph-catenary data, historical track data, and historical tunnel data. Each transaction dataset contains L transactions corresponding to the L data types.

[0145] The second determination module is configured to determine multiple frequent itemsets based on the support and support threshold of each transaction.

[0146] The generation module is configured to generate multiple frequent sub-itemsets corresponding to each frequent itemset based on each transaction in each frequent itemset.

[0147] The calculation module is configured to calculate the lift of each frequent subset and determine the relationships between transactions in the frequent subset based on the lift.

[0148] The second construction module is configured to build a predetermined graph of each transaction based on the relationship between each transaction.

[0149] According to embodiments of this disclosure, the L types of data include: historical running stability data, which includes at least one or a combination of the following data types: target train smoothness detection data, instability detection data, vibration detection data, and idling coasting data; historical pantograph-catenary data, which includes at least one or a combination of the following data types: target train catenary pressure data, catenary current data, pantograph-catenary contact pressure data, track arcing rate data, and pantograph-catenary foreign object data; historical track data, which includes at least one or a combination of the following data types: rail profile data, rail corrugation data, turnout abnormal state data, track gauge detection data, and rail surface state data during idling coasting; and historical tunnel data, which includes at least one or a combination of the following data types: tunnel foreign object intrusion data, tunnel clearance data, and tunnel video anomaly data.

[0150] According to embodiments of this disclosure, the target graph neural network model is trained by the following method, including a third acquisition module and an optimization module.

[0151] The third acquisition module is configured to acquire a predetermined map and historical fault diagnosis information within a historical time period.

[0152] The optimization module is configured to use a predetermined graph as historical sample data and historical fault diagnosis information as labels to optimize the initial graph neural network model and obtain the target graph neural network model.

[0153] According to embodiments of this disclosure, the fault determination method based on the rail-tunnel system further includes a third determination module.

[0154] The third determination module is configured to determine the target fault point from potential fault points based on the second real-time operating status information, second real-time operating environment information and fault occurrence probability of multiple reference trains in the same area, so as to eliminate the fault at the target fault point in advance.

[0155] According to embodiments of this disclosure, the potential fault information further includes the potential fault type and the expected occurrence time of the potential fault. The fault determination method based on the rail-tunnel system further includes a fourth determination module and a fifth determination module.

[0156] The fourth determination module is configured to determine the severity and scope of impact of a potential fault based on the potential fault location, the potential fault type, and the expected occurrence time of the potential fault.

[0157] The fifth module is configured to determine the priority and handling strategy for potential faults based on their severity and scope of impact.

[0158] Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure, or at least part of the functions of any one or more of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be implemented by dividing them into multiple modules. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as hardware circuitry, such as Field Programmable Gate Arrays (FPGAs), Programmable Logic Arrays (PLAs), Systems-on-Chip, Systems-on-Substrate, Systems-on-Package, Application-Specific Integrated Circuits (ASICs), or implemented in hardware or firmware by any other reasonable means of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0159] For example, any plurality of the first acquisition module 410A, the first determination module 420A, and the output module 430A can be combined into one module / unit / subunit, or any one of these modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least part of the functionality of one or more of these modules / units / subunits can be combined with at least part of the functionality of other modules / units / subunits and implemented in one module / unit / subunit. According to embodiments of this disclosure, at least one of the first acquisition module 410A, the first determination module 420A, and the output module 430A can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the first acquisition module 410A, the first determination module 420A, and the output module 430A may be implemented at least partially as a computer program module, which can perform corresponding functions when the computer program module is run.

[0160] The embodiments of this disclosure provide a train operation scheduling method based on a track-tunnel system, based on the potential fault information obtained by the fault determination method described above. The method includes: acquiring potential fault information detected when a reference train is traveling on an initial line, wherein the potential fault information includes one or more of the following: catenary fault information, track fault information, and tunnel fault information related to the operating environment of the initial line; acquiring first operating status information, first mileage information, first maintenance information, and multiple first waiting-to-run line information of a target train within a first historical time period, wherein the target train is a train planned to travel on the initial line within a target preset time period; generating first health status information of the target train based on the first operating status information, first mileage information, and first maintenance information; and generating a first operating strategy for the target train based on the first health status information and potential fault information, to control the target train to continue operating on the initial line or to schedule the target train to run on a first alternative line other than the initial line from the multiple first waiting-to-run line information.

[0161] Figure 5 schematically illustrates an exemplary system architecture for applying a train operation scheduling method based on a network track tunnel system according to embodiments of the present disclosure. It should be noted that Figure 5 is merely an example of a system architecture to which embodiments of the present disclosure can be applied, to help those skilled in the art understand the technical content of the present disclosure, but does not imply that embodiments of the present disclosure cannot be used in other devices, systems, environments, or scenarios.

[0162] As shown in Figure 5, the system architecture 100B according to this embodiment includes a track-tunnel system 110B, a control center 120B, a train monitoring system 130B, and a positioning and synchronization system 140B.

[0163] The track-rail-tunnel system 110B includes a pantograph-catenary detection system 111B, a track detection system 112B, a tunnel detection system 113B, and a main unit 114B, as well as other related subsystems or systems.

[0164] The pantograph-catenary inspection system 111B is configured to acquire catenary parameter data, pantograph status data, and pantograph-catenary relationship data collected by pantograph-catenary inspection equipment installed on the top of the train. Examples include conductor pull-out value detection, pantograph-catenary arc detection, temperature detection, and hard spot detection.

[0165] The track inspection system 112B is configured to acquire track geometric parameter data, track component status data, and train operating environment data collected by track inspection equipment installed on the bottom of the train. The track inspection system 112B also includes: a track condition inspection device, a track geometric parameter detection device, and a rail profile detection device (not shown in the figure). The track condition inspection device can acquire track image information acquired at high speed by the track inspection equipment, and extract damage information such as rail surface scratches, fish scale damage, spalling, wave wear, ballast bed cracks, and missing fasteners from the images. It uses a high-precision camera to dynamically acquire data on the track surface, and then uses digital image processing technology to automatically identify defects in fasteners, rail surfaces, and ballast bed, and transmit faults in real time. The track geometric parameter detection device can establish a navigation coordinate system to achieve relative positioning and attitude measurement of the vehicle. Simultaneously, combined with a vibration compensation device, it achieves vibration compensation and track gauge measurement; combined with a positioning synchronization system, it achieves precise positioning synchronization. Finally, it outputs track geometric parameter information such as track gauge, superelevation, track direction, and elevation in real time. The rail profile detection device can perform real-time, dynamic, and high-precision detection of the entire rail cross-section and wear. It automatically establishes a standard rail profile, matches and analyzes each collected profile with the standard profile, and outputs vertical and horizontal wear results in real time.

[0166] The tunnel inspection system 113B is configured to acquire tunnel clearance data and tunnel structural status data collected by tunnel inspection equipment installed at the front of a train or in a tunnel. Specifically, a high-speed laser scanner can be used to quickly scan the tunnel surface, acquiring three-dimensional information of the tunnel cross-section and presenting the spatial morphology of the tunnel. The high-speed acquired tunnel cross-section data is compared with a preset subway clearance model (which specifies the dimensions and other standards that subway tunnel spaces must meet) to determine whether the actual tunnel space meets the standards, obtaining the tunnel clearance inspection results and checking for problems such as encroachment. The acquired two-dimensional images are used to identify surface defects in the tunnel, such as cracks and spalling, to assess the tunnel's health status.

[0167] The main unit 114B is usually installed in the car cabinet or under the seat, and is configured to collect, organize, analyze, determine faults and provide early warnings, and store data collected by the pantograph-catenary detection system 111B, track detection system 112B, and tunnel detection system 113B.

[0168] Control center 120B includes a track maintenance management system 121B, a train dispatching system 122B, other related subsystems, and terminal equipment (not shown in the diagram). The track maintenance management system 121B is configured to receive track and tunnel data transmitted from the track and tunnel system 110B, and to classify, store, and analyze this data for later retrieval. The train dispatching system 122B is configured to monitor train positions and status in real time, formulate and adjust train operation strategies, and coordinate train passing and avoidance maneuvers. The terminal equipment provides a visual interface to display data processing and analysis results, early warning information, etc., to track maintenance personnel, enabling them to monitor vehicle and environmental information in real time and develop contingency plans for potential faults.

[0169] The Train Monitoring System 130B monitors the train's operating status in real time, such as speed and direction of travel; the status of train equipment, such as traction system, braking system, bogies, and doors; and also includes: train number information, train mileage, train fault diagnosis information, train signaling equipment PHM (Prognostics and Health Management) data and related diagnostic information. Among them, the train PHM diagnostic data can be used to guide daily train inspection activities.

[0170] The positioning and synchronization system 140B obtains train location information by using radio frequency tags installed at specific locations on the track or by satellite positioning.

[0171] Figure 6 schematically illustrates a flowchart of a train operation scheduling method according to an embodiment of the present disclosure.

[0172] As shown in Figure 6, the train operation scheduling method 200B includes operations S210B to S240B, which are performed by the train scheduling system.

[0173] During operation S210B, potential fault information detected when the reference train was traveling on the initial track is acquired. This potential fault information includes one or more of the following: overhead contact line fault information, track fault information, and tunnel fault information, all related to the operating environment of the initial track.

[0174] According to embodiments of this disclosure, the reference train can be a train currently traveling on the initial track. Potential fault information is information predicted by the track maintenance management system 121B based on real-time operating environment information detected by the track-tunnel system 110B, indicating potential faults. Examples include: contact wire wear, loose suspension devices, track gauge changes, rail cracks, and tunnel water leakage.

[0175] During operation S220B, the system acquires the target train's first operating status information, first mileage information, first maintenance information, and multiple first lines to be operated within a first historical time period. The target train is the train scheduled to travel on the initial line within a target preset time period.

[0176] According to embodiments of this disclosure, when potential fault points exist in the operating environment related to the initial line, a target train is determined based on the initial line. Specifically, a train scheduled to travel on the initial line within a preset target time period (e.g., within 5 hours) is determined as the target train. Within the current time frame, the target train can be in operation or in a planned operation state (stationary state).

[0177] According to embodiments of this disclosure, the first historical time period can be any time period prior to the current moment, for example, the time elapsed since the last completion of train maintenance. The first operating status information characterizes the real-time operating status of the target train within the first historical time period, such as the train's speed, acceleration, and the operating status of various systems on the train (e.g., operating parameters of the traction and braking systems). The first mileage information is the mileage traveled by the target train within the first historical time period, which can be used to assess the wear and tear of train components. The first maintenance information characterizes the maintenance records of the target train within the first historical time period, such as maintenance time, components maintained, and maintenance methods. Multiple first routes to be operated indicate that, in addition to the initial route, the target train can also select other routes to operate on.

[0178] When operating S230B, the first health status information of the target train is generated based on the first operating status information, the first operating mileage information, and the first maintenance information.

[0179] In operation S240B, based on the first health status information and potential fault information, a first operating strategy is generated for the target train. This strategy controls the target train to continue operating on the initial line or to be dispatched to a first alternative line from among multiple first alternative lines to be operated.

[0180] According to embodiments of this disclosure, the first operating strategy may be: if the target train is in a good first health state and the potential fault of the initial line will not seriously affect the train operation, the first operating strategy may be to control the target train to continue running on the initial line; conversely, if the target train has a health problem, or the potential fault of the initial line may threaten the safety of the target train operation, the first operating strategy may be to dispatch the target train to a first alternative line other than the initial line from multiple first lines to be operated, thereby ensuring the safety and reliability of the train operation.

[0181] According to embodiments of this disclosure, the distribution and status of problem areas are intuitively displayed through a visual interface, enabling maintenance personnel to quickly and accurately locate and handle potential faults, thereby improving the safety and reliability of the rail transit system.

[0182] For example, the initial line is a section of a high-speed rail line from point A to point B. Some time ago, or currently, a reference train traveling on this line detected a minor water leak in a tunnel section. The assessment concluded that the leak would not significantly impact train operation in the short term. The target train is a high-speed rail train scheduled to travel from point A to point B. In the past month (the first historical time period), the target train's initial operational status information shows that its traction and braking systems, among other critical systems, are operating stably with all parameters normal; the initial mileage information indicates it has traveled 50,000 kilometers, and the wear of various components is within normal limits; the initial maintenance information indicates that the train underwent a comprehensive inspection and maintenance last month, replacing some easily damaged parts, and the vehicle is currently in good condition. Since the target train's initial health status is good, and the potential fault in the initial line (minor water leak) will not significantly affect train operation, the initial operational strategy is to control the target train to continue operating on the initial line as planned, but to slow down when passing through the leaking tunnel.

[0183] According to embodiments of this disclosure, by acquiring potential fault information detected by a reference train on the initial line, such as potential fault information of the overhead contact system, track, and tunnel, potential fault hazards in the line can be predicted in advance. By comprehensively analyzing information such as the target vehicle's operating status, mileage, and maintenance, health status information is generated, thereby allowing for real-time adjustments to the operating strategy of the target train scheduled to travel to that section of the line. Specifically, by acquiring predicted potential fault information, the approach shifts from passively responding to faults to proactively preventing risks, thus minimizing adjustments to the operating strategy. Specifically, since there may be correlations between different lines, for example, rescheduling a target train to another line may require multiple trains to modify their current routes. This disclosure, based on the health status of the target vehicle and predicted potential faults (e.g., based on predicted potential fault information), allows maintenance personnel to investigate or repair potential fault points within a reasonable timeframe (e.g., before the target train reaches the faulty section), proactively avoiding potential fault risks. This prevents adjustments to multiple lines due to local line adjustments, ensuring the stability of railway transportation. On the other hand, when the target train has health problems or faults that cannot be repaired within a limited time, and the potential faults on the line threaten operational safety, dispatching the target train to an alternative line can prevent the train from running on dangerous lines, reduce the probability of accidents such as derailment and collisions, and ensure the safety of the train and passengers.

[0184] According to an embodiment of this disclosure, a first health status information of the target train is generated based on the first operating status information, the first operating mileage information, and the first maintenance information, including steps 11 to 12.

[0185] Step 11: Based on the first operating status information, the first operating mileage information, and the first maintenance information, generate the target train's energy consumption information and loss information.

[0186] Step 12: Based on energy consumption information and loss information, generate the first health status information of the target train.

[0187] According to embodiments of this disclosure, the energy consumption information of the target train is used to characterize the energy consumption of the target train during a first historical time period. For example, during acceleration, a large amount of energy is required to overcome the train's stationary inertia and initial running resistance, resulting in typically high energy consumption. During constant-speed travel, energy consumption is relatively stable, mainly used to maintain the train's speed and overcome various running resistances. Since the operating conditions of trains are similar and regular to some extent, it is possible to predict the energy consumption of the target train continuing to travel on the initial route based on the current energy consumption information. For example, if historical data shows the energy consumption of the train under similar route conditions (such as gradient, curve radius, etc.), a more accurate prediction of the energy consumption on the initial route can be made.

[0188] According to embodiments of this disclosure, the loss information of the target train is used to characterize the wear, aging, and performance degradation of various components and systems of the train during a first historical time period due to various effects during operation (such as mechanical friction, electrical aging, environmental corrosion, etc.). From the perspective of mechanical components, long-term contact and friction between wheels and rails leads to wear on the wheel treads, affecting the train's operational stability and safety. During long-term operation, the suspension system, due to the continuous bearing of the train's weight and vibration impacts, will gradually experience fatigue wear on components such as springs and shock absorbers, reducing the performance of the suspension system. Since the train's loss information can reflect changes in the performance indicators of some key systems, such as a decrease in braking force of the braking system and unstable power output of the traction system, monitoring the loss information helps prevent the target train from traveling on complex lines under conditions of significant loss.

[0189] According to embodiments of this disclosure, after obtaining energy consumption information and loss information, both are comprehensively analyzed. Abnormal energy consumption is often related to the performance status of train equipment; excessively high or low energy consumption may indicate equipment malfunction or decreased efficiency. Loss information directly reflects the wear and tear of various train components; severe wear may affect the normal function of components and even cause safety hazards. By establishing an evaluation model, the energy consumption information and loss information are quantified and compared with preset health standards or historical data to generate the first health status information of the target train.

[0190] According to embodiments of this disclosure, energy consumption information and loss information of a target train are generated based on first operating status information, first operating mileage information, and first maintenance information. This includes: inputting the first operating status information, first operating mileage information, and first maintenance information into a trained first target model, and outputting energy consumption information and loss information; wherein, the first target model is obtained by training a first initial model using the historical energy consumption information and historical loss information of sample trains as labels, and utilizing the historical operating status information, historical operating mileage information, and historical maintenance information of sample trains.

[0191] The historical energy consumption and wear information of the sample trains are associated with their historical operating mileage, historical operating environment, and historical maintenance records. The first target model can be trained using machine learning algorithms such as neural networks, random forests, and support vector machines. The algorithm used can be adjusted according to specific circumstances, and this disclosure does not impose any limitations on it.

[0192] For example, a neural network algorithm is used. The initial model is trained by inputting historical operating status, mileage, and maintenance information of each sample component from the sample train, and using historical energy consumption and wear information of the sample components as output labels. When the error between the output energy consumption or wear value and the true value exceeds a predetermined threshold, the model parameters are adjusted to ensure the model can fit the relationship between the input data and the output value as accurately as possible. This adjustment continues until the error between the predicted energy consumption or wear value and the true label value meets the predetermined threshold, at which point the model parameter tuning terminates. After training and validation using a large amount of historical data from sample components, the first target model is obtained.

[0193] In addition, before using data for model training and calculation, in order to improve the accuracy of model calculation, data cleaning and standardization operations can be performed on the original data. For example, abnormal data and duplicate data can be deleted, data can be normalized, and data can be converted into values ​​between [0, 1].

[0194] According to embodiments of this disclosure, by inputting the historical operating status information, historical mileage information, and historical maintenance information of the sample train into the trained first target model, the energy consumption value or loss value of the target train can be calculated accurately and quickly, reducing errors caused by manually set parameters, improving the accuracy of energy consumption value or loss value calculation, effectively assessing the health status of the target train, and thereby generating corresponding train operation strategies to improve the efficiency of train operation scheduling.

[0195] According to embodiments of this disclosure, a first health status information of a target train is generated based on energy consumption information and loss information, including: inputting energy consumption information, loss information and first running mileage information into a trained second target model, and outputting the health status information of the target train; wherein, the second target model is obtained by training a second initial model using historical energy consumption information, historical loss information and historical running mileage information, with the historical failure rate of each sample train as a label.

[0196] According to embodiments of this disclosure, similar to the first target model, the second target model can also be trained using machine learning algorithms such as neural networks, random forests, and support vector machines. The algorithm used can be adjusted according to specific circumstances, and this disclosure does not limit it. The specific training process and data preprocessing process are similar to those of the first target model and can be adjusted according to actual conditions, and will not be elaborated here.

[0197] According to embodiments of this disclosure, by constructing a target model of train health status information and inputting the energy consumption information, loss information and first running mileage information of the same train into a trained second target model, the health status information of the target train can be determined quickly and accurately, reducing the error in health status assessment caused by experience error, improving the accuracy of train health status assessment, and thereby generating corresponding train operation strategies to improve the efficiency of train operation scheduling.

[0198] According to embodiments of this disclosure, a positioning synchronization system is used to locate the train in order to calculate the first mileage information of the target train. Positioning can be performed using methods such as Global Positioning System (GPS), Differential Global Positioning System (DSGPS), Inertial Navigation System (INS), wheel and axle encoders, wireless communication positioning, track circuits, radio frequency identification (RFID), visual positioning, and lidar.

[0199] Among these methods, the Global Positioning System (GPS) uses satellite-provided location information to determine a vehicle's exact location. GPS offers high accuracy, reaching within meters in open areas, but signal loss may occur in tunnels or under viaducts. Differential GPS (GPS) uses ground-based base stations to transmit correction signals, improving GPS accuracy and making it suitable for applications requiring higher positioning precision. Inertial Navigation Systems (INS) use sensors such as accelerometers and gyroscopes to measure the vehicle's motion and calculate its current position based on initial position data. INS can serve as a supplementary positioning method when GPS signals are weak or absent. Wheel and axle encoders monitor wheel and axle rotations to calculate the distance traveled. This method is low-cost but highly susceptible to variations in wheel diameter and slippage. Track circuitry utilizes the electrical characteristics of the track to determine the train's position. When the train passes through a specific area, it alters the current state of that area, thus achieving positioning. Wireless communication positioning, including technologies like Wi-Fi and Bluetooth, estimates the distance between the vehicle and a known location by measuring signal strength or time-of-flight, thereby deducing the vehicle's position. This method is suitable for precise positioning within stations or specific areas. Radio Frequency Identification (RFID) involves placing RFID tags along the track. When a train passes, the tag information is read to determine the train's position. This method is suitable for precise positioning at fixed points. Visual positioning uses onboard cameras to capture markers or specific patterns along the trackside, and image processing technology determines the train's position. This method requires good environmental conditions, such as sufficient lighting and clearly visible markers. LiDAR (Light Detection and Ranging) constructs a 3D model of the surrounding environment by emitting a laser beam and measuring the reflection time, thus determining the vehicle's position. This method is suitable for precise positioning in complex environments.

[0200] According to embodiments of this disclosure, during the positioning process, the detection control box is responsible for acquiring encoder or speed sensor signals. The encoder is installed on parts such as wheel axles, and calculates the travel distance by detecting the number of axle rotations, thus obtaining the mileage. Speed ​​sensor signals can be acquired through speed sensor devices, and the travel distance can also be calculated by integrating speed over time. The detection control box processes these signals to calculate the mileage. Due to factors such as equipment errors and wheel wear causing changes in wheel diameter, the mileage calculated solely from encoder or speed sensor signals may contain some errors. To ensure the accuracy of the mileage data, synchronous calibration of the mileage is required to make the mileage recorded by the system as consistent as possible with the actual travel distance, which helps improve positioning accuracy.

[0201] According to embodiments of this disclosure, when the system can interconnect with a vehicle signaling system, the vehicle signaling system provides two pieces of information. First, the vehicle's real-time mileage, which is relatively accurate mileage data obtained by the vehicle signaling system through its own precise calculations or by fusion of multiple sensors. Second, the vehicle clock signal, used to unify the time standard. Using the real-time mileage provided by the vehicle signaling system, the system undergoes mileage synchronization correction, adjusting the mileage data calculated by the detection control box to match the mileage provided by the vehicle signaling system. Simultaneously, the system's time is synchronized using the vehicle clock signal, ensuring that the time of each part within the system remains synchronized with the time of the vehicle signaling system, thereby improving the accuracy of the system's mileage data and time.

[0202] According to embodiments of this disclosure, when the system cannot interconnect with the vehicle signaling system, an alternative method is used for mileage synchronization correction and clock timing. An RFID reader is installed, and RFID tags are pre-positioned on the track. Each RFID tag stores information such as its precise location on the track. When a vehicle passes over an RFID tag, the RFID reader reads the tag signal. Based on the read tag location information and previously recorded driving data, the system performs mileage synchronization correction, adjusting the recorded mileage data. Simultaneously, this process is used to provide unified clock timing for all subsystems, ensuring time consistency across all subsystems and guaranteeing the accuracy and coordination of the entire system's time and mileage data.

[0203] Figure 7 schematically illustrates the working principle of a positioning synchronization system according to an embodiment of the present disclosure.

[0204] As shown in Figure 7, the positioning and synchronization system 140B includes an RFID tag reader 141B, a positioning and synchronization server 142B, a speed sensor 143B, and a detection control box 144B.

[0205] RFID reader 141B is installed at specific locations along the track. When a train passes these locations, the reader reads the RFID tags on the train to obtain its location information, such as determining if the train has passed a station or a specific track section. The RFID reader 141B then sends the obtained train location information to a location synchronization server.

[0206] The positioning synchronization server 142B receives train location information sent from the RFID tag reader 141B.

[0207] Speed ​​sensor 143B is installed on the train and generates mileage pulse signals in real time as the train travels. For example, speed sensor 143B generates a pulse signal for every meter traveled, and these pulse signals are continuously output.

[0208] The detection control box 144B collects pulse signals sent by the speed sensor 143B and calculates the train's real-time mileage based on the number and frequency of the pulse signals. For example, if 100 pulse signals are received, and each pulse represents 1 meter, the train has traveled 100 meters. After calculating the mileage, the detection control box 144B outputs a TTL level trigger pulse to transmit information such as mileage calculation completion. Simultaneously, it provides a DC24V power interface to power related equipment, ensuring stable signal transmission and normal equipment operation. The detection control box 144B sends the calculated mileage information to the positioning synchronization server 142B, enabling the positioning synchronization server 142B to obtain the train's mileage data.

[0209] As shown in Figure 7, the positioning synchronization server 142B receives mileage information from the detection control box 144B and train position information from the RFID reader 141B. It then integrates these two types of information. The mileage of the train at a certain location (determined by the RFID reader 141B) (calculated by the detection control box 144B) is correlated and calibrated to ensure the accuracy and consistency of the position and mileage information. The mileage information, after being integrated and processed by the positioning synchronization server 142B, is distributed to other systems via a switch, such as the track maintenance management system 121B in Figure 5, or directly sent to the train dispatching system 122B.

[0210] According to embodiments of this disclosure, the train dispatching system 122B receives mileage information, location information, and health status information of the target vehicle, and generates an operation strategy for the target train to improve transportation efficiency and resource utilization.

[0211] According to embodiments of this disclosure, the positioning synchronization server 142B also includes a mileage calibration module. The mileage calibration module acquires the position signal transmitted by the RFID reader 141B and calculates the calibration mileage using the position information according to the mileage calibration protocol. By comparing the continuous mileage with that of the speed sensor 143B, it performs real-time calibration on other detection systems within the mileage synchronization network, correcting continuous mileage errors. The continuous mileage of the speed sensor 143B may exhibit a "cliff effect" (sudden discontinuous jumps in mileage data). The calibration mileage obtained from the RFID reader 141B and the vehicle mileage signal is used to stretch (supplement data when mileage data decreases briefly) or compress (delete data when mileage data increases briefly) the continuous mileage to ultimately obtain accurate mileage, ensuring that the mileage data accurately reflects the actual train operation.

[0212] According to embodiments of this disclosure, by fusing information from multiple sources such as speed sensors, RFID tags, and vehicle signals, calculation deviations are corrected, enabling mileage data to accurately reflect the actual travel distance, improving positioning accuracy, and providing reliable mileage information for train operation control, scheduling, maintenance management, and other systems, thus ensuring the stable operation of each system.

[0213] According to embodiments of this disclosure, potential fault information includes potential fault points, and health status information includes healthy and unhealthy states. Based on the first health status information and the potential fault information, a first operating strategy is generated for the target train to control the target train to continue operating on the initial line or to dispatch the target train to a first alternative line other than the initial line from a plurality of first alternative lines to be operated, including steps 21 to 23. The potential fault point's handling level is determined based on the potential fault information, where the handling level includes urgent and non-urgent.

[0214] Step 21: If the target train's first health status information is healthy and the pending level is non-emergency, control the target train to continue running on the initial line.

[0215] Step 22: If the target train's first health status information is healthy and the pending level is emergency, the target train will be dispatched to one of the first alternative lines.

[0216] Step 23: If the first health status information of the target train is unhealthy, the target train will be dispatched to one of the first alternative lines.

[0217] According to embodiments of this disclosure, controlling the target train to continue running on the initial track can generate a speed reduction strategy. Specifically, the speed reduction magnitude is determined based on the severity of the track condition. For example, mild degradation might only require a 10% speed reduction, while severe degradation might require a speed reduction of more than 30%. The specific area for speed reduction is defined; it can be an entire section or a specific track segment. The time window for speed reduction is determined; for example, normal speed can be restored during nighttime maintenance, while speed reduction is required during the day.

[0218] According to embodiments of this disclosure, the speed reduction strategy is communicated to the dispatching and command center, which then coordinates the operation plans of all trains. The speed reduction information is sent to relevant stations and train drivers through the train dispatching system to ensure they understand and execute the speed reduction instruction. The speed reduction information is also disseminated through official websites, apps, and other channels to inform passengers of potential delays.

[0219] According to embodiments of this disclosure, potential faults can be further classified into minor, moderate, and severe levels based on the information. Minor faults: Continue monitoring operation while ensuring safety, increase the frequency of inspections, and record the fault details. Moderate faults: Limit speed operation in the faulty section, adjust the timetable to reduce train density, and arrange temporary repairs during off-peak hours. Severe faults: Immediately suspend train service in the affected section, organize a professional team for emergency repairs, promptly inform passengers and provide alternative transportation suggestions.

[0220] According to embodiments of this disclosure, when generating an operation strategy, it is also necessary to assess the impact of the fault on train operation (delays, cancellations, speed limits), passengers (number of passengers affected, peak passenger flow), transfers (transfers at hub stations), and determine the track sections directly affected by the fault point, taking into account the impact on adjacent sections and power supply zones.

[0221] According to embodiments of this disclosure, for trains that are healthy and have no urgent matters, they are allowed to continue operating on their initial lines, reducing unnecessary scheduling, maintaining efficient operation of the main lines, ensuring overall operational order, and improving transportation efficiency. Conversely, when a train is unhealthy or has urgent matters to attend to, it is scheduled to an alternative line, preventing faulty trains from causing more serious accidents on the main lines, avoiding disruption to other normal train operations, and ensuring the operational safety of the entire rail transit network.

[0222] According to an embodiment of this disclosure, when the first operating strategy is to dispatch the target train to the first alternative line, the method further includes steps 31 to 32.

[0223] Step 31: Obtain the operating environment monitoring information of each first alternative line.

[0224] Step 32: Based on the first health status information of the target train and the operating environment monitoring information of each first alternative line, generate a scheduling plan for the target train to schedule the target train to run on one of the first alternative lines.

[0225] According to embodiments of this disclosure, in addition to considering alternative routes, it is also necessary to consider whether the alternative routes currently meet the conditions for train operation. By comprehensively considering the health status of the target train and the operating environment of the alternative routes, it is possible to avoid dispatching trains to routes unsuitable for their current conditions, reducing operational risks and ensuring the safety of trains and passengers. At the same time, it also avoids situations where some alternative routes are overused while others are idle due to blind dispatching, thus improving overall operational efficiency.

[0226] According to embodiments of this disclosure, the train operation scheduling method further includes steps 41 to 43.

[0227] Step 41: Obtain the second operating status information, second operating mileage information, second maintenance information, and multiple second lines to be operated information of the reference train within the second historical time period.

[0228] Step 42: Based on the second operating status information, the second operating mileage information, and the second maintenance information, generate the second health status information of the reference train.

[0229] Step 43: Based on the second health status information and potential fault information, generate a second operating strategy for the reference train to control the reference train to continue running on the initial line or to dispatch the reference train to run on a second alternative line other than the initial line from multiple second alternative lines to be run information.

[0230] According to embodiments of this disclosure, when a potential fault exists on the initial line of the reference train, it is necessary not only to update the operating strategy of subsequent target trains planned to travel on the initial line, but also to prioritize resolving the current second operating strategy of the reference train. Similar to the above implementation, second health status information of the reference train is generated using second operating status information, second mileage information, second maintenance information, and multiple second alternative lines to be operated within a second historical time period. A second operating strategy for the reference train is generated by combining the second health status information and potential fault information. If the reference train is in good health and has no potential faults or the potential faults do not affect its operation on the initial line, the strategy may be to allow the reference train to continue operating on the initial line. If the train has certain health problems or potential faults, and it is determined that the initial line needs to be avoided to ensure operational safety or facilitate maintenance, the train will be dispatched to a second alternative line other than the initial line from the multiple second alternative lines to be operated information.

[0231] Based on the above-described train operation scheduling method, this application also provides a train operation scheduling device. The device will be described in detail below with reference to Figure 8.

[0232] Figure 8 schematically illustrates a structural block diagram of a train operation scheduling device according to an embodiment of this application.

[0233] As shown in Figure 8, the train operation scheduling device 400B of this embodiment includes a first acquisition module 410B, a second acquisition module 420B, a first generation module 430B, and a second generation module 440B.

[0234] The first acquisition module 410B is configured to acquire potential fault information detected when the reference train is traveling on the initial track. The potential fault information includes one or more of the following: overhead contact line fault information, track fault information, and tunnel fault information related to the operating environment of the initial track. In one embodiment, the first acquisition module 410B can be used to perform the operation S210B described above, which will not be repeated here.

[0235] The second acquisition module 420B is configured to acquire the target train's first operating status information, first operating mileage information, first maintenance information, and multiple first lines to be operated information within a first historical time period. The target train is a train that is scheduled to travel on the initial line within a target preset time period. In one embodiment, the second acquisition module 420B can be used to execute the operation S220B described above, which will not be repeated here.

[0236] The first generation module 430B is configured to generate first health status information of the target train based on first operating status information, first operating mileage information, and first maintenance information. In one embodiment, the first generation module 430B can be used to perform the operation S230B described above, which will not be repeated here.

[0237] The second generation module 440B is configured to generate a first operating strategy for the target train based on the first health status information and potential fault information, so as to control the target train to continue running on the initial line or to dispatch the target train to a first alternative line other than the initial line from a plurality of first lines to be run. In one embodiment, the second generation module 440B can be used to execute the operation S240B described above, which will not be repeated here.

[0238] According to an embodiment of this disclosure, the first generation module 430B includes: a first generation submodule and a second generation submodule.

[0239] The first generation submodule is configured to generate target train energy consumption information and loss information based on the first operating status information, the first operating mileage information and the first maintenance information; the second generation submodule is configured to generate target train first health status information based on energy consumption information and loss information.

[0240] According to an embodiment of this disclosure, the first generation submodule includes: a first output unit configured to input first operating status information, first operating mileage information and first maintenance information into a trained first target model, and output energy consumption information and loss information; wherein, the first target model is obtained by training a first initial model using the historical energy consumption information and historical loss information of the sample train as labels, and using the historical operating status information, historical operating mileage information and historical maintenance information of the sample train.

[0241] According to an embodiment of this disclosure, the second generation submodule includes: a second output unit configured to input energy consumption information, loss information and first running mileage information into a trained second target model, and output health status information of the target train; wherein, the second target model is obtained by training a second initial model using historical energy consumption information, historical loss information and historical running mileage information, with the historical failure rate of each sample train as a label.

[0242] According to embodiments of this disclosure, potential fault information includes potential fault points, health status information includes healthy and unhealthy, and the second generation module 440B includes: a determination submodule, a control submodule, a first scheduling submodule, and a second scheduling submodule.

[0243] The system comprises three submodules: a determination submodule, configured to determine the processing level of potential fault points based on potential fault information, where processing levels include urgent and non-urgent; a control submodule, configured to control the target train to continue running on the initial line when the target train's first health status information is healthy and the processing level is non-urgent; a first scheduling submodule, configured to schedule the target train to one of the first alternative lines when the target train's first health status information is healthy and the processing level is urgent; and a second scheduling submodule, configured to schedule the target train to one of the first alternative lines when the target train's first health status information is unhealthy.

[0244] According to embodiments of this disclosure, when the first operating strategy is to dispatch the target train to the first alternative line, the train operation scheduling device further includes: a third acquisition module and a third generation module.

[0245] The third acquisition module is configured to acquire the operating environment monitoring information of each of the first candidate lines. The third generation module is configured to generate a scheduling plan for the target train based on the target train's first health status information and the operating environment monitoring information of each of the first candidate lines, so as to schedule the target train to run on one of the first candidate lines.

[0246] According to embodiments of this disclosure, the train operation scheduling device further includes a fourth acquisition module and a fourth generation module.

[0247] The fourth acquisition module is configured to acquire the second operating status information, second mileage information, second maintenance information, and multiple second alternative routes for the reference train within the second historical time period. The fourth generation module is configured to generate second health status information for the reference train based on the second operating status information, second mileage information, and second maintenance information; and to generate a second operating strategy for the reference train based on the second health status information and potential fault information, to control the reference train to continue operating on the initial route or to dispatch the reference train to a second alternative route other than the initial route from among the multiple second alternative routes.

[0248] Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure, or at least part of the functions of any one or more of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be implemented by dividing them into multiple modules. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as hardware circuitry, such as Field Programmable Gate Arrays (FPGAs), Programmable Logic Arrays (PLAs), Systems-on-Chip, Systems-on-Substrate, Systems-on-Package, Application-Specific Integrated Circuits (ASICs), or implemented in hardware or firmware by any other reasonable means of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0249] For example, any plurality of the first acquisition module 410B, the second acquisition module 420B, the first generation module 430B, and the second generation module 440B can be combined into one module / unit / subunit, or any one of these modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least part of the functionality of one or more of these modules / units / subunits can be combined with at least part of the functionality of other modules / units / subunits and implemented in one module / unit / subunit. According to embodiments of this disclosure, at least one of the first acquisition module 410B, the second acquisition module 420B, the first generation module 430B, and the second generation module 440B can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the first acquisition module 410B, the second acquisition module 420B, the first generation module 430B, and the second generation module 440B can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0250] Overhaul refers to the process of disassembling, cleaning, inspecting, flaw-detecting, and repairing important components of a vehicle when its mileage or operating time reaches specified values. It also involves comprehensive testing, adjustments, and trials to restore the vehicle's overall performance and meet regulatory requirements and quality acceptance standards. Daily train inspection refers to a comprehensive technical inspection of the train conducted daily, before departure after train formation or before disassembly at a station. This inspection ensures the train is in good technical condition, promptly identifies and addresses potential faults, and guarantees safe train operation.

[0251] In traditional electromechanical system integration, information is independent and control is decentralized between the rail vehicle system and various ground systems, resulting in low efficiency in collaborative operation and scheduling. For example, the train dispatcher, depot dispatcher, and yard dispatcher each perform their own duties; vehicle operation faults require drivers to communicate with depot and yard dispatchers via telephone or fax data; daily vehicle inspections are performed manually; there is no trackside inspection equipment, and there is no data exchange between daily inspections and overhauls, nor between vehicle operation data and maintenance data; feedback on train status information, maintenance information, and track status information is not timely. Therefore, assessing the health status of trains and the status of train operation, and scheduling trains based on these assessments, consumes significant manpower and resources and is time-consuming, resulting in low efficiency in train operation scheduling, fault diagnosis, and maintenance.

[0252] Therefore, this embodiment of the disclosure also integrates the network track tunnel system with the information of rail vehicles, and uses the potential fault information identified by the network track tunnel system as the monitoring information of the line to be operated, and participates in the train operation scheduling of the vehicle-ground integrated system.

[0253] In view of this, the embodiments of this disclosure, through a vehicle-to-ground integrated system, can promptly acquire status monitoring information, mileage information, maintenance record information, historical operating environment information of multiple candidate trains, and monitoring information of multiple lines to be operated from various components configured on the train. Based on the above information, corresponding health status information of each candidate train and status information of each line to be operated are generated. Through the generated health status information and status information of the lines to be operated, a reasonable train operation scheduling plan is finally generated efficiently to schedule each candidate train to run on each line to be operated. The above-mentioned process of generating the operation scheduling plan utilizes data interaction between the vehicle and ground daily inspection systems to more rationally and efficiently evaluate the health status of the train and the status of the train's operating lines, effectively shortening the time for formulating the operation scheduling plan and improving the efficiency of train operation scheduling.

[0254] The embodiments disclosed herein relate to vehicle systems and ground maintenance systems within a vehicle-ground integrated system. Centered on vehicle operation and maintenance, they utilize advanced technologies such as artificial intelligence, industrial internet, internet of things, operations research optimization, and 3D visualization. Through real-time perception of vehicle operation and maintenance status, combined with optimization of maintenance processes for key vehicle components, the system reorganizes vehicle maintenance production models and re-plans depot layout structures. This achieves seamless integration of depot data with vehicle data and information flows, enabling vehicle data to guide depot maintenance and operation, while maintenance data supports full lifecycle vehicle management. Ultimately, this generates highly operable and reliable train operation scheduling methods, improving train operation scheduling efficiency and providing support for the standardization and specifications of new digital depot construction.

[0255] Figure 9 schematically illustrates an exemplary system architecture for applying a train operation scheduling method based on a vehicle-ground integrated system according to embodiments of the present disclosure. It should be noted that Figure 9 is merely an example of a system architecture applicable to embodiments of the present disclosure, intended to help those skilled in the art understand the technical content of the disclosure, and does not imply that embodiments of the present disclosure cannot be used in other devices, systems, environments, or scenarios.

[0256] As shown in Figure 9, the system architecture 100C according to this embodiment includes a vehicle system 110C, a ground signal system 120C, a ground inspection and maintenance system 130C, and a ground application service platform 140C.

[0257] The vehicle system 110C includes the onboard network subsystem 111C, the onboard signaling equipment 112C, and other related subsystems. During train operation, the data provided by the train includes, but is not limited to, the following: train operating mode, such as traction mode, braking mode, coasting mode, traction or braking level information, train mileage, train fault diagnosis information, train signaling equipment PHM (Prognostics and Health Management) data and related diagnostic information. Among these, the train PHM diagnostic data can be used to guide daily train inspection activities.

[0258] The aforementioned data can be transmitted by the vehicle to the ATS (Automatic Train Supervision) subsystem of the ground signaling system 120C via wireless communication channels, where information is overlaid on the acquired vehicle data. The data overlaid by the ground signaling system 120C includes, but is not limited to, the following: train number information, train position information, and train speed information.

[0259] The ground signaling system 120C then transmits data to the ground maintenance system 130C. The data provided by the ground maintenance system 130C to the ground application service platform 140C includes, but is not limited to, the following: daily inspection and diagnostic information, component replacement information, and repair schedule information. This data is then transmitted back to the ground application service platform 140C.

[0260] In addition, the data transmission path can be changed to transmit vehicle data and signal data separately, which are then aggregated in the ground data center before being sent to the ground inspection and maintenance system. Whether or not a ground data center is set up does not affect the operation of the entire system.

[0261] The ground inspection and maintenance system 130C adds a new daily train inspection system 131C and an intelligent control platform 132C. The daily train inspection system 131C refers to the system that performs technical inspections on trains every day. The new components can be configured according to different application scenarios. For example, the new daily train inspection system 131C can add a 360° trackside inspection of the car body and an intelligent inspection robot for the wheelset pantograph. This disclosure does not impose any restrictions on this.

[0262] According to embodiments of this disclosure, information such as the status monitoring and maintenance records of each component can be obtained through the novel daily inspection system 131C. Figure 10 schematically illustrates the working principle of the novel daily inspection system according to an embodiment of this disclosure.

[0263] As shown in Figure 10, when a vehicle enters the throat area, a 360° exterior image detection of the vehicle can be triggered to identify the vehicle number, measure speed, and collect images of the roof, sides, and undercarriage. This identifies daily inspection items and faults on the vehicle's exterior, which are then remotely viewed by the inspection team. After the vehicle enters the depot, daily inspection operations are triggered. The inspection team issues automated inspection commands to collect and identify images of daily inspection items on the undercarriage running gear, and remotely views the images and faults via a handheld terminal. After reviewing the fault images, the inspection team performs maintenance checks on the daily inspection faults and remaining items. After the daily inspection is completed, the maintenance information is fed back to the ground application service platform, and the train completes its maintenance status record.

[0264] Among the aforementioned interactive data, vehicle operating mode, traction level information, braking level information, vehicle position, and vehicle speed play crucial roles. Taking vehicle status records related to overhauls as an example, when equipment undergoes multiple overhauls, vehicle position and / or speed can be used to determine whether the overhaul occurred in the same area and / or at the same speed level, facilitating subsequent equipment improvements and fault analysis. This data can be used for fault coupling analysis, which involves transmitting overhaul data, along with corresponding vehicle operating data, vehicle health data, signal data, and daily inspection data, to the overhaul equipment supplier for data coupling judgment and analysis at the time of the fault.

[0265] Regarding train maintenance malfunctions, the vehicle system can perform self-checks based on the fault detection logic of each system. When a fault condition threshold is met, the system automatically reports the fault, locates the specific equipment component, and feeds the fault information back to the ground for maintenance and repair. The ground system can monitor the status information of various train equipment based on image recognition. When a fault condition threshold is met, the system automatically reports the fault, locates the specific equipment component, and feeds it back for maintenance and repair. For example, if a scratch on a high-voltage equipment between the train and the ground exceeds 5mm, the system will report a fault. Furthermore, since train fault diagnosis data may not be accurate, and the fault may not be caused by a train malfunction, but rather indirectly by a track fault at the train's location, the fault will not occur as long as the train is not traveling on that section of track. For example, if a train reports a "traction converter grounding fault," and inspection reveals it was caused by a foreign object impact, the train's location at the time the fault was reported can guide track inspection to prevent recurrence without requiring a major overhaul.

[0266] It should be understood that the number of systems, subsystems, and devices in Figure 9 is merely illustrative. The number of systems, subsystems, and devices can be increased or decreased according to the actual application scenario.

[0267] Figure 11 schematically illustrates a flowchart of a train operation scheduling method based on a vehicle-ground integrated system according to an embodiment of the present disclosure.

[0268] As shown in Figure 11, the method 300C includes operations S310C to S340C.

[0269] By operating S310C, the system acquires status monitoring information of each component configured on multiple candidate trains, mileage information of each component, maintenance record information of each component, historical operating environment information of multiple candidate trains, and monitoring information of multiple lines to be operated.

[0270] When operating S320C, health status information of each candidate train is generated based on the status monitoring information of each component, the operating mileage information of each component, the maintenance record information of each component, and the historical operating environment information of each candidate train.

[0271] When operating the S330C, the status information of each line to be operated is generated based on the monitoring information of each line to be operated.

[0272] When operating S340C, based on the health status information of each candidate train and the status information of each line to be operated, an operation scheduling plan is generated for each candidate train to be scheduled to run on each line to be operated.

[0273] According to embodiments of this disclosure, the status monitoring information of various components configured on the train includes status monitoring information of components in various subsystems of the vehicle system. This can include status monitoring information of components such as the traction system, power supply system, and vehicle-to-ground communication equipment, to achieve status diagnosis of the components configured on the train. Specifically, this can include status monitoring information of components such as the train bogie, car body, doors, brakes, traction, air conditioning, broadcasting, pyrotechnics, auxiliary systems, passenger information system, vehicle-to-ground wireless communication equipment, the vehicle (Train Control and Management System) network itself, and communication equipment between the vehicle TCMS and other system networks. For example, status monitoring information might include "cracks exist in the bogie of train X," "charger malfunctions in train X," or "traction equipment is normal for train X." The status monitoring information of each component can be obtained during vehicle self-inspection or through ground system diagnostics, and the two can exchange information via vehicle-to-ground communication.

[0274] The mileage information of each component can be extracted from the train's mileage information, representing the number of kilometers each component has traveled. For example, the P traction motor has traveled 75,200 kilometers. The mileage information of each component includes the mileage when the train is carrying passengers and when the train is empty.

[0275] The maintenance records for each component include both maintenance information and upkeep information. Both maintenance and upkeep records can be used as a basis for determining the remaining service life of a component. For example, the maintenance record information might be: "Due to cracks, the bogie of Train X underwent a major overhaul and replacement on October 25, 2018. The estimated service life of the replaced bogie is 10 years."

[0276] Train operating environment information includes track conditions, train operating mode, traction, and braking level information. For example, if the train's historical operating environment information is "the train has passed through 10 curves and 5 steep slopes," then the traction wheels and bogies may be severely worn.

[0277] According to embodiments of this disclosure, by utilizing the status monitoring information, mileage information, maintenance record information, and historical operating environment information of each candidate train obtained above, a health status assessment of the train is achieved through multi-parameter fusion calculation.

[0278] The health status information of each candidate train can be used to assign corresponding weight coefficients to the condition monitoring data, mileage data, maintenance record data, and operating environment data of different components through the analytic hierarchy process and expert experience method. Finally, the data are summed to obtain the health status score of each component. The health status score of the component with the lowest score is used as the total health status score of the train, so as to quantify the health status information of each candidate train.

[0279] For example, based on expert experience, the weights for traction motor condition monitoring, mileage, maintenance records, and operating environment are 0.8, 0.5, 0.6, and 0.4, respectively. The processed values ​​for these data are 90, 70, 60, and 85, respectively. Therefore, the sum of these weights yields a health status score of 177 for each component. Similarly, the weights for bogie condition monitoring, mileage, maintenance records, and operating environment are 0.9, 0.6, 0.8, and 0.8, respectively. The processed values ​​for these data are 79, 68, 80, and 90, respectively. Therefore, the sum of these weights yields a health status score of 248 for each component. In this case, the lower score of 177 is chosen as the overall health status score for the train.

[0280] The health status information of each candidate train can also be evaluated by assessing the status monitoring data of different components, mileage data, maintenance record data, and operating environment data. The lowest level among these assessments is taken as the train's health status level to evaluate the train's health status.

[0281] For example, after evaluating the status monitoring data, mileage data, maintenance record data, and operating environment data of the traction motor and bogie, the traction motor and bogie are classified as Level 1 and Level 2. Level 1 indicates that the component is in poor condition, and Level 2 indicates that the component is in moderate condition. Therefore, the lower level, Level 1, is selected as the health status evaluation of the train.

[0282] The above assessment results can also be presented to operations and maintenance personnel through visualization charts such as bar charts, line charts, and statistical tables.

[0283] The monitoring information of the lines to be operated can be obtained by the on-board track inspection system after assessing the condition of the ground tracks. Based on the monitoring information of each line to be operated, the status information of each line to be operated is generated, which can be used to characterize whether the lines to be operated can operate normally. The monitoring information of the lines to be operated includes the track condition information, power usage information, and other information of the lines to be operated.

[0284] According to embodiments of this disclosure, an operation scheduling plan for each candidate train is generated by matching the health status information of each candidate train with the status information of the line to be operated. The matching of the train's health status information and the line's status information can be based on a generated matching list of the corresponding train's health status and the line's status, a mapping table between the two, or matching rules developed through expert experience.

[0285] For example, in the matching list, if a train's health status is level a, it can be matched with the operating line statuses of levels a, b, and c. If the health status information of train X is "Good health status, level a", and the status information of operating line Y is "Train line status level b", then the health status information of train X can be matched with the status information of operating line Y, and train X can be scheduled to run according to operating line Y.

[0286] According to embodiments of this disclosure, the vehicle-to-ground integrated system can promptly acquire status monitoring information, mileage information, maintenance record information, historical operating environment information of multiple candidate trains, and monitoring information of multiple waiting lines for operation from various components configured on the train. Based on the above information, health status information of each candidate train and status information of each waiting line for operation are generated accordingly. Using the generated health status information and status information of the waiting lines, a reasonable train operation scheduling plan is ultimately generated efficiently to schedule each candidate train to operate on each waiting line. The above-mentioned process of generating the operation scheduling plan utilizes data interaction between the vehicle and ground-based daily inspection systems to more rationally and efficiently evaluate the health status of the train and the status of the train's operating lines, effectively shortening the time required to formulate the operation scheduling plan and improving the efficiency of train operation scheduling.

[0287] According to embodiments of this disclosure, health status information of each candidate train is generated based on the status monitoring information of each component, the operating mileage information of each component, the maintenance record information of each component, and the historical operating environment information of each candidate train. This includes: extracting the historical operating environment information of each component from the historical operating environment information of each candidate train; generating the remaining lifespan information of each component based on the operating mileage information, the historical operating environment information, and the maintenance record information of each component; and generating the health status information of each candidate train based on the remaining lifespan information and the status monitoring information of each component of the same candidate train.

[0288] Figure 12 schematically illustrates a process for generating health status information according to an embodiment of the present disclosure.

[0289] As shown in Figure 12, based on the train's historical operating environment information 421C, the component's historical operating environment information 422C can be extracted. This information includes the frequency and duration of traction and braking, the degree of impact or wear, and moisture levels. Both the component's historical operating environment information 422C and its mileage information 423C reflect its service life to some extent. The number of repairs and maintenance performed also affect the assessment of its remaining service life. Therefore, the component's remaining service life information 425C can be determined based on its historical operating environment information 422C, mileage information 423C, and maintenance record information 424C.

[0290] For example, a corresponding lifespan conversion factor can be determined based on the component's historical operating environment information, mileage information, and maintenance record information, thereby generating a depletion lifespan value. Using the factory-estimated lifespan as the initial value, the depletion lifespan value obtained above is subtracted to finally obtain the component's remaining lifespan value.

[0291] The condition monitoring information 426C of the component can be converted into numerical values ​​through data standardization processing, thereby quantifying the condition monitoring information of the component and obtaining the condition monitoring value of the component.

[0292] Based on the remaining lifespan information 425C of the components and the condition monitoring information 426C of the components, the health status information 427C of the train can be obtained. Specifically, the health status value of each component can be calculated by multiplying the remaining lifespan value of the component and the condition monitoring value of the component, and the minimum health status value of all components of the train is taken as the health status value of the train.

[0293] According to embodiments of this disclosure, historical operating environment information of components is extracted from the historical operating environment information of the train. Based on the component's mileage information, historical operating environment information, and maintenance record information, the remaining lifespan information of each component is obtained. Based on the remaining lifespan information and condition monitoring information, the train's health status information is generated. By using the remaining lifespan of components as an indicator to assess the train's health status, a more comprehensive and accurate evaluation and diagnosis of the train based on component usage is achieved. This prevents unpredictable component failures, generates corresponding train operation scheduling strategies, and improves the safety of vehicle operation scheduling.

[0294] According to embodiments of this disclosure, based on the mileage information, historical operating environment information, and maintenance record information of each component, the remaining lifespan information of each component is generated. This includes: inputting the mileage information, historical operating environment information, and maintenance record information of each component into a trained first target model, and outputting the remaining lifespan information of each component. The first target model is obtained by training a first initial model using the historical remaining lifespan of the sample components as labels, and utilizing historical mileage, historical operating environment, and historical maintenance records.

[0295] The historical remaining life of the sample components refers to the historical remaining life of each historically scrapped component at different historical moments. The magnitude of the historical remaining life is related to historical operating mileage, historical operating environment, and historical maintenance records.

[0296] The first target model can be trained using machine learning algorithms such as neural networks, random forests, and support vector machines. The algorithm used can be adjusted according to specific circumstances, and this disclosure does not limit it.

[0297] For example, a neural network algorithm is used. The historical mileage, historical operating environment, and historical maintenance records of each sample component are input, and the historical remaining lifespan of the sample component is used as the output label to train the first initial model. When the error between the output remaining lifespan value and the true value exceeds a predetermined threshold, the model parameters are adjusted to ensure that the model can fit the relationship between the input data and the output value as accurately as possible. This adjustment continues until the error between the predicted remaining lifespan value and the true label value meets the predetermined threshold, at which point the model parameter tuning terminates. After training and validation with a large amount of historical data from sample components, the first target model is obtained.

[0298] In addition, before using data for model training and calculation, in order to improve the accuracy of model calculation, data cleaning and standardization operations can be performed on the original data. For example, abnormal data and duplicate data can be deleted, data can be normalized, and data can be converted into values ​​between [0, 1].

[0299] According to embodiments of this disclosure, by inputting the mileage information, historical operating environment information, and maintenance record information of each component into a trained first target model, the remaining life information of each component can be calculated accurately and quickly, reducing errors caused by manually set parameters, improving the accuracy of remaining life calculation, effectively assessing the health status of the train, and thereby generating corresponding train operation scheduling strategies to improve the efficiency of train operation scheduling.

[0300] According to embodiments of this disclosure, health status information of each candidate train is generated based on the remaining lifespan information and status monitoring information of each component of the same candidate train. This includes: inputting the remaining lifespan information and status monitoring information of each component of the same candidate train into a trained second target model, and outputting the health status information of each candidate train respectively; wherein, the second target model is obtained by training a second initial model using the historical failure rate of each sample train as a label and the historical remaining lifespan and historical monitoring status of the components of each sample train.

[0301] According to embodiments of this disclosure, the historical failure rate of the sample train is the percentage of failures of each historically scrapped component at different historical times out of the total number of uses. For example, the failure rate of the traction motor from time t to time t of scrapping. n The failure rate at any given time is 1%.

[0302] Similar to the first objective model, the second objective model can also be trained using machine learning algorithms such as neural networks, random forests, and support vector machines. The algorithm used can be adjusted according to specific circumstances, and this disclosure does not impose any limitations on it. The specific training process and data preprocessing process are similar to those of the first objective model and can be adjusted according to actual conditions, and will not be elaborated here.

[0303] According to embodiments of this disclosure, by constructing a target model of train health status information and inputting the remaining lifespan information and status monitoring information of each component of the same candidate train into the trained second target model, the health status information of each candidate train can be determined quickly and accurately, reducing the error in health status assessment caused by experience error, improving the accuracy of train health status assessment, and thereby generating a corresponding train operation scheduling strategy to improve the efficiency of train operation scheduling.

[0304] According to embodiments of this disclosure, the monitoring information of the line to be operated includes the monitoring information of the overhead contact system, track monitoring information, and tunnel monitoring information corresponding to the line to be operated; based on the monitoring information of each line to be operated, the state information of each line to be operated is generated, including: inputting the monitoring information of each overhead contact system, each track monitoring information, and each tunnel monitoring information corresponding to each line to be operated into a trained third target model to generate the state information of each line to be operated; wherein, the third target model is obtained by training a third initial model using the state of the sample operating line as a label and utilizing the historical monitoring information of the overhead contact system, the historical monitoring information of the track, and the historical monitoring information of the tunnel corresponding to the sample operating line.

[0305] According to embodiments of this disclosure, the monitoring information for each overhead contact line includes the contact line voltage and current information. Track monitoring information includes the degree of track wear, track geometry, track spacing, and other information. Tunnel monitoring information includes tunnel environmental information, tunnel length, wind speed within the tunnel, and airflow pressure, and other information.

[0306] The status of the sample operating lines includes the status information of historical operating lines under the same operating route and speed. By inputting the monitoring information of each overhead contact line, each track monitoring information, and each tunnel monitoring information corresponding to each line to be operated into the trained third target model, the status information of each line to be operated can be generated.

[0307] Similar to the first and second objective models, the third objective model can also be trained using machine learning algorithms such as neural networks, random forests, and support vector machines. The algorithm used can be adjusted according to specific circumstances, and this disclosure does not impose any limitations on it. The specific training process and data preprocessing process are similar to the objective models mentioned above and can be adjusted according to actual conditions, so they will not be elaborated here.

[0308] According to embodiments of this disclosure, a third target model is trained using historical monitoring information of the overhead contact system, track, and tunnel corresponding to the sample running line. The third target model is then used to calculate and analyze the status information of the running line, which can reduce errors caused by manual assessment and judgment, improve the accuracy of train running line status assessment, and generate corresponding train operation scheduling strategies to improve the efficiency of train operation scheduling.

[0309] Figure 13 schematically illustrates a flowchart of a train operation scheduling method based on a vehicle-ground integrated system according to an embodiment of the present disclosure.

[0310] As shown in Figure 13, the method 500C includes operations S541C to S545C.

[0311] In operation S541C, the initial matching result is obtained by matching the health status information of each candidate train with the status information of each line to be operated.

[0312] When operating S542C, the initial matching results are verified based on the remaining lifespan information of each component in each candidate train, and the verification results are obtained.

[0313] In operation S543C, determine whether the verification result is passed. If the result is yes, execute operation S544C; if the result is no, execute operation S545C.

[0314] When operating the S544C, a runtime scheduling plan is generated based on the initial matching results.

[0315] When operating the S545C, adjust the initial matching results to generate an operation scheduling plan.

[0316] After sorting the health status information of each candidate train and the status information of each line to be operated, the train is matched. Specifically, the matching can be done through the corresponding mapping relationship between the health status information and the status information of the operating line. For example, the mapping relationship is "if the health status information is greater than 80 points, it is matched with the line with good operating status; if the health status information is less than 75 points, the train will not depart". If the generated health status information a is 85 points and the operating line status information b is good, then a and b can be matched.

[0317] The initial matching results may have the following issues: If the health status determined by each component's state happens to meet the threshold, but the matched operating line status is below average, the component may not meet the requirements of the current operating line. For example, component x may have a long service life and severe wear, but due to its good historical operating environment, infrequent maintenance, and frequent upkeep, its health status score is high. However, the remaining lifespan of this component may be insufficient for operating lines with many curves and braking. Therefore, it is necessary to verify the initial matching results using the component's remaining lifespan information to reduce the risk of operational scheduling and improve the safety of train operation scheduling.

[0318] During verification, verification can be performed according to verification rules. For example, a verification rule could be that, under normal operating conditions, the remaining service life of a component is not less than one year. When the verification result is successful, an operation scheduling plan can be generated based on the initial matching result. When the verification result is unsuccessful, the matching result needs to be adjusted to meet the verification rules. After passing the verification, a corresponding operation scheduling plan is generated based on the adjusted matching result.

[0319] According to embodiments of this disclosure, matching the train's health status information with the status information of the line to be operated can quickly yield matching results. In addition, the initial matching results are verified by using the remaining lifespan information of components. Taking into account potential safety hazards during the initial matching, a more reasonable and safer operation scheduling plan is generated.

[0320] Based on the above-described train operation scheduling method based on a vehicle-ground integrated system, this disclosure also provides a train operation scheduling device based on a vehicle-ground integrated system. The device will be described in detail below with reference to Figure 14.

[0321] Figure 14 schematically illustrates a structural block diagram of a train operation scheduling device based on a vehicle-ground integrated system according to an embodiment of the present disclosure.

[0322] As shown in Figure 14, the train operation scheduling device 600C based on the vehicle-ground integrated system in this embodiment includes an acquisition module 610C, a first generation module 620C, a second generation module 630C, and a third generation module 640C.

[0323] The acquisition module 610C is configured to acquire status monitoring information of each component configured on multiple candidate trains, mileage information of each component, maintenance record information of each component, historical operating environment information of multiple candidate trains, and monitoring information of multiple lines to be operated. In one embodiment, the acquisition module 610C may be configured to execute the operation S310C described above, which will not be repeated here.

[0324] The first generation module 620C is configured to generate health status information for each candidate train based on the status monitoring information of each component, the operating mileage information of each component, the maintenance record information of each component, and the historical operating environment information of each candidate train. In one embodiment, the first generation module 620C may be configured to execute the operation S320C described above, which will not be repeated here.

[0325] The second generation module 630C is configured to generate status information for each line to be operated based on monitoring information of each line. In one embodiment, the second generation module 630C can be used to execute the operation S330C described above, which will not be repeated here.

[0326] The third generation module 640C is configured to generate an operation scheduling plan for each candidate train based on the health status information of each candidate train and the status information of each line to be operated, so as to schedule each candidate train to run on each line to be operated. In one embodiment, the third generation module 640C can be used to execute the operation S340C described above, which will not be repeated here.

[0327] According to embodiments of this disclosure, the first generation module 620C includes an extraction submodule, a remaining lifespan information generation submodule, and a health status information generation submodule. The extraction submodule is configured to extract historical operating environment information of each component from the historical operating environment information of each candidate train. The remaining lifespan information generation submodule is configured to generate remaining lifespan information for each component based on its mileage information, historical operating environment information, and maintenance record information. The health status information generation submodule generates health status information for each candidate train based on the remaining lifespan information and status monitoring information of each component within the same candidate train.

[0328] According to an embodiment of this disclosure, the remaining lifespan information generation submodule includes a first target model application unit, configured to input the operating mileage information of each component, the historical operating environment information of each component, and the maintenance record information of each component into a trained first target model, and output the remaining lifespan information of each component respectively; wherein, the first target model is obtained by training a first initial model using the historical remaining lifespan of the sample component as a label, and utilizing the historical operating mileage, historical operating environment, and historical maintenance record.

[0329] According to an embodiment of this disclosure, the health status information generation submodule includes a second target model application unit, configured to input the remaining life information and status monitoring information of each component of the same candidate train into the trained second target model, and output the health status information of each candidate train respectively; wherein, the second target model is obtained by training the second initial model using the historical failure rate of each sample train as a label and the historical remaining life and historical monitoring status of the components of each sample train.

[0330] According to embodiments of this disclosure, the monitoring information for the lines to be operated includes monitoring information of the overhead contact system, track monitoring information, and tunnel monitoring information corresponding to the lines to be operated. The second generation module 630C includes a third target model application submodule, configured to input the monitoring information of each overhead contact system, each track monitoring information, and each tunnel monitoring information corresponding to each line to be operated into a trained third target model to generate state information for each line to be operated; wherein, the third target model is obtained by training a third initial model using the state of a sample operating line as a label, and utilizing historical monitoring information of the overhead contact system, the track, and the tunnel corresponding to the sample operating line.

[0331] According to embodiments of this disclosure, the third generation module 640C includes a matching submodule, a verification submodule, and a response submodule. The matching submodule is configured to perform matching based on the health status information of each candidate train and the status information of each line to be operated, to obtain an initial matching result. The verification submodule is configured to verify the initial matching result based on the remaining lifespan information of each component in each candidate train, to obtain a verification result. The response submodule is configured to generate an operation scheduling plan based on the initial matching result in response to a successful verification result.

[0332] Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure, or at least part of the functions of any one or more of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be implemented by dividing them into multiple modules. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as hardware circuitry, such as Field Programmable Gate Arrays (FPGAs), Programmable Logic Arrays (PLAs), Systems-on-Chip, Systems-on-Substrate, Systems-on-Package, Application-Specific Integrated Circuits (ASICs), or implemented in hardware or firmware by any other reasonable means of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0333] For example, any plurality of the acquisition module 610C, the first generation module 620C, the second generation module 630C, and the third generation module 640C can be combined into one module / unit / subunit, or any one of these modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least part of the functionality of one or more of these modules / units / subunits can be combined with at least part of the functionality of other modules / units / subunits and implemented in one module / unit / subunit. According to embodiments of the present disclosure, at least one of the acquisition module 610C, the first generation module 620C, the second generation module 630C, and the third generation module 640C can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the acquisition module 610C, the first generation module 620C, the second generation module 630C, and the third generation module 640C can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0334] Embodiments of this disclosure also include a train, comprising an acquisition module and a generation module. The acquisition module is configured to acquire the operation scheduling plan obtained by the above method. The generation module is configured to generate operation instructions based on the operation scheduling plan to control the train to operate according to the operation scheduling plan.

[0335] In traditional electromechanical system integration, the design of various disciplines such as vehicles, signaling, power supply, and depots is independent, information is isolated, and control is decentralized. In particular, the coordination efficiency between ground systems and vehicles is low, making it difficult to achieve effective breakthroughs in improving the performance of the entire system.

[0336] In terms of rail vehicle maintenance, the current conventional urban rail ground maintenance system operates with separate functions for dispatching, depot dispatching, and yard dispatching. When a vehicle malfunctions, the driver must communicate with the depot dispatcher via telephone or fax to complete the repair and dispatching. This manual dispatching method is inefficient. Furthermore, there is no data exchange between the vehicle maintenance data and operational data of each maintenance system. For example, there is no data exchange between the overhaul system and the vehicles, resulting in information silos between the maintenance systems and the vehicles. This leads to untimely feedback on maintenance status, conflicts over site occupancy, and a year-on-year increase in maintenance costs.

[0337] In recent years, some progress has been made in intelligent control systems for rail vehicles. For example, the exploration of the DCC (Dynamic Control Center) integrated management model for subways and the application of intelligent control systems in some subway depots; the use of automated production lines and logistics equipment for the maintenance of certain components in some train depots, etc., have improved the efficiency of rail vehicle operation scheduling and management to a certain extent. Therefore, given the current problems between the vehicle fault repair and maintenance system and the vehicles themselves, it is necessary to propose a new intelligent control method to achieve timely feedback and repair of vehicle faults, thereby improving the efficiency of vehicle repair and maintenance.

[0338] Therefore, in this embodiment of the present disclosure, the potential fault information obtained based on the track-tunnel system described above can be used as part of the initial fault diagnosis information, and the vehicle-ground integrated system can be configured to combine the historical operating status of the train and the initial fault diagnosis information to determine the fault to be repaired.

[0339] In view of this, the embodiments of this disclosure utilize a vehicle-to-ground integrated system to promptly acquire vehicle operating status information, initial fault diagnosis information, and ground maintenance operation status information. Based on this information, the information of components to be repaired can be determined, and the time required for the vehicle to reach the ground maintenance site can be calculated. Based on the aforementioned information of components to be repaired, the time required to reach the ground maintenance site, and the ground maintenance operation status information, a vehicle maintenance scheduling strategy is generated. The vehicle operates according to the aforementioned scheduling strategy, waiting for repair. During this process, vehicle-to-ground communication between the vehicle and the ground maintenance system enables data exchange and sharing, establishing a seamless data flow and information flow between the ground maintenance system and the vehicle. When a vehicle malfunctions, the system can promptly combine vehicle operating status information, fault information, ground maintenance records, and ground maintenance operation status information to generate a reasonable vehicle scheduling and maintenance strategy. This at least partially solves the problem of untimely scheduling at ground maintenance sites affecting vehicle fault repair efficiency, improves vehicle fault repair efficiency, shortens the vehicle fault repair waiting period, and ensures that repaired vehicles can promptly participate in vehicle operation scheduling, reducing the time cost of fault repair.

[0340] Figure 15 schematically illustrates a vehicle-to-ground integrated system applicable to a vehicle maintenance scheduling method according to an embodiment of the present disclosure. It should be noted that Figure 15 is merely an example of a system architecture to which embodiments of the present disclosure can be applied, to help those skilled in the art understand the technical content of the present disclosure, but does not imply that embodiments of the present disclosure cannot be used in other devices, systems, environments, or scenarios.

[0341] The aforementioned vehicle-ground integrated system 100D, which can be applied to vehicle maintenance and scheduling methods, includes a vehicle-ground collaborative maintenance system. Specifically, it mainly consists of a vehicle part and a ground part. The vehicle part mainly involves the TCMS and PHM subsystems of the vehicle system 140D, as well as related traction and braking subsystems. The ground part mainly includes the signal system 130D, the ground maintenance system 110D, and the corresponding ground application service platform 120D.

[0342] During vehicle operation, information such as mileage, PHM (Prognostics and Health Management), and major fault information can be transmitted in real time to the ground application service platform 120D via the vehicle-to-ground communication channel. The signal system 130D overlays the vehicle location information and forwards the vehicle operation status data on the line to the ground maintenance system 110D via the ground application service platform 120D. At the same time, maintenance data is obtained in real time through ground communication connection, and maintenance data, fault handling data, and other information are fed back to the vehicle system 140D, and each vehicle is managed with a unique file.

[0343] Compared to traditional ground maintenance systems, the ground maintenance system 110D disclosed herein adds an intelligent control platform 111D and an intelligent maintenance system 112D, wherein the intelligent maintenance system includes an intelligent frame overhaul system 112D-1.

[0344] The intelligent control platform 111D can generate vehicle maintenance and depot entry / exit plans based on vehicle operating status and fault diagnosis information, using data such as vehicle mileage. Vehicles are then parked on appropriate tracks according to the return-to-depot and maintenance plans, awaiting maintenance work. Fault repair plans are initiated by maintenance personnel who obtain work orders, assess the depot equipment status, and collect necessary materials before commencing repair work. Upon completion of maintenance, work orders are submitted, and dispatchers arrange vehicle departure based on the completion status and vehicle exit plan. Simultaneously, maintenance information is fed back to the vehicle system 140D to record train maintenance status. Data mining and analysis are performed on fault and maintenance data generated during the maintenance process, and data is integrated with the vehicle system 140D's PHM subsystem and the ground application service platform 120D to provide fault model-based maintenance guidance.

[0345] Figure 16 schematically illustrates the working principle of the intelligent control platform according to an embodiment of the present disclosure.

[0346] As shown in Figure 16, the intelligent control platform includes intelligent management of DCC depot scheduling operations, information-based control of vehicle maintenance, status and assessment of depot equipment assets, safety assurance and linkage of maintenance operations, and maintenance data application and analysis. The intelligent management module in DCC depot scheduling operations sends management plans to the information-based control module for vehicle maintenance. The information-based control module for vehicle maintenance and the status and assessment module for depot equipment assets send maintenance data and equipment data to the maintenance data application and analysis module, respectively, realizing intelligent control of ground maintenance. Maintenance personnel perform relevant operations corresponding to the above modules. For example, maintenance team members and quality inspection personnel interact with the information-based control module for vehicle maintenance to retrieve and fill out work orders. The ground maintenance system communicates and interacts with the ground application service platform, the ground signal system, and the vehicle system to achieve vehicle-ground collaborative intelligent maintenance and scheduling management.

[0347] Taking the intelligent overhaul system 112D-1 as an example, the operation process of the intelligent maintenance system is described. After acquiring train mileage and fault diagnosis information, the intelligent overhaul system 112D-1 automatically arranges overhaul vehicles to appropriate maintenance tracks according to the prepared vehicle overhaul plan. The overhaul team's maintenance personnel are responsible for all maintenance stages, including vehicle decoupling, component disassembly, cleaning, assembly, testing, and vehicle final assembly and debugging, according to the maintenance plan and work orders. Material management personnel promptly distribute materials through material warehousing management. During the maintenance process, the overhaul information system provides dynamic guidance on maintenance processes and collects maintenance data, forming electronic maintenance records for components. At the same time, maintenance information is fed back to the vehicle system 140D to complete the train maintenance status record and update relevant key vehicle information, such as wheelset automatic correction information.

[0348] Figure 17 schematically illustrates the working principle of the intelligent frame overhaul system according to an embodiment of the present disclosure.

[0349] As shown in Figure 17, the intelligent overhaul system includes information-based management and control of the overhaul process, automated inspection and repair of key vehicle components, and material warehousing management and distribution. The information-based management and control module for the overhaul process is configured to distribute inspection process standards to the automated inspection and repair module for key vehicle components. The automated inspection and repair module for key vehicle components interacts with the information-based management and control module for the overhaul process. The material warehousing management and distribution module automatically delivers components to the automated inspection and repair module for key vehicle components. Maintenance personnel perform corresponding operations with the modules mentioned above; for example, warehouse logistics managers perform material management and selection operations, coordinating with the material warehousing management and distribution system. The ground maintenance system communicates and interacts with the ground application service platform, ground signal system, and vehicle system to achieve vehicle-ground collaborative intelligent vehicle maintenance and scheduling management.

[0350] It should be understood that the number of system modules and their connections in Figures 15-17 are merely illustrative. Depending on implementation needs, there can be any number and function of system modules.

[0351] Figure 18 schematically illustrates a flowchart of a vehicle maintenance scheduling method based on a vehicle-body integrated system according to an embodiment of the present disclosure.

[0352] As shown in Figure 18, the method 400D includes operations S410D to S450D.

[0353] The S410D is used to acquire operating status information, initial fault diagnosis information, and ground maintenance operation status information for multiple vehicles.

[0354] When operating S420D, the historical mileage and initial fault diagnosis information of each vehicle are analyzed to determine the information of each component to be repaired.

[0355] When operating the S430D, the travel time required for each vehicle to reach the ground maintenance and repair site is calculated based on the operating location and speed of each vehicle.

[0356] When operating the S440D, maintenance scheduling strategy information for multiple vehicles is generated based on information about each component to be inspected, each travel duration, and ground maintenance operation status.

[0357] When operating the S450D, based on the maintenance dispatch strategy information, each vehicle is dispatched to drive into the ground maintenance site according to its corresponding track information to wait for maintenance.

[0358] According to embodiments of this disclosure, the operating status information includes the vehicle's operating location, operating speed, and historical mileage, where the historical mileage can reflect the possible causes of vehicle malfunctions and the service life of various vehicle components. Initial fault diagnosis information includes the faulty vehicle's number and fault location, used to characterize the initial scope of the vehicle malfunction; for example, the initial fault diagnosis information might be: "Train X1025 traction motor malfunction, unable to operate normally." The operating status information and initial fault diagnosis information can be obtained through communication with the vehicle system.

[0359] The ground maintenance operation status information indicates the specific details of the ground maintenance and can be obtained through various ground subsystems.

[0360] According to embodiments of this disclosure, the component to be inspected can be any component of the vehicle that has malfunctioned, such as the traction motor, temperature sensor, speed sensor, bogie, critical body components, bearings, etc., without limitation. The information of the component to be inspected includes the vehicle number to which the component belongs, the name and model of the component, the cause of the component's malfunction, and the maintenance plan for the component.

[0361] Based on the vehicle's historical mileage and initial fault diagnosis information, the information of the parts to be repaired can be determined. For example, based on the information that "X1025 vehicle has reached 1.7 million kilometers, traction motor bearing failure, traction operation failure", the vehicle's mileage meets the requirements of the overhaul plan, and the traction motor bearing needs to be replaced. Therefore, the information of the parts to be repaired can be determined as "the traction motor bearing P1213 of X1025 vehicle is severely worn and needs to be overhauled and the bearing replaced".

[0362] According to embodiments of this disclosure, after determining the information of the component to be repaired, the location of the ground repair and maintenance site that the vehicle needs to enter can be determined according to the repair and maintenance plan. Based on the vehicle's operating location and speed, the travel time required for the vehicle to reach the designated ground repair and maintenance site can be calculated.

[0363] Based on information about the components to be inspected, travel time, and ground maintenance operation status, a maintenance dispatch strategy can be generated tailored to the vehicle's specific situation. This strategy specifies the track information the vehicle will be on while waiting for maintenance at the ground maintenance depot, including the track number and position. It also indicates the time the vehicle will enter the designated track to wait for maintenance. According to this strategy, the vehicle is dispatched to the ground maintenance depot to await repair, following the corresponding track information.

[0364] According to embodiments of this disclosure, a vehicle-to-ground integrated system can promptly acquire vehicle operating status information, initial fault diagnosis information, and ground maintenance operation status information. Based on this information, the information of components to be repaired can be determined, and the time required for the vehicle to reach the ground maintenance location can be calculated. Based on the aforementioned information of components to be repaired, the time required to reach the ground maintenance location, and the ground maintenance operation status information, a vehicle maintenance scheduling strategy is generated. The vehicle operates according to the aforementioned scheduling strategy, awaiting maintenance. During this process, vehicle-to-ground communication between the vehicle and the ground maintenance system enables data exchange and sharing, establishing a seamless data flow and information flow between the ground maintenance system and the vehicle. When a vehicle malfunctions, a reasonable vehicle scheduling and maintenance strategy can be generated by promptly combining vehicle operating status information, fault information, ground maintenance records, and ground maintenance operation status information. This achieves accuracy in vehicle maintenance, improves the efficiency of vehicle fault repair, and reduces maintenance costs.

[0365] According to embodiments of this disclosure, maintenance scheduling strategy information for multiple vehicles is generated based on information about each component to be inspected, each travel duration, and ground maintenance operation status information. This includes: determining each waiting duration and each maintenance duration based on information about each component to be inspected and ground maintenance operation status information; wherein each waiting duration represents the time required for each vehicle to wait for maintenance; each maintenance duration represents the time required to maintain each vehicle; and maintenance scheduling strategy information is generated based on each waiting duration, each maintenance duration, and each travel duration.

[0366] The total time spent on the vehicle maintenance process includes the time it takes for the vehicle to travel from its current location to the ground maintenance site, the waiting time, and the time required for the maintenance itself. Therefore, it is necessary to comprehensively consider the waiting time, maintenance time, and travel time to develop a reasonable maintenance scheduling strategy.

[0367] Figure 19 schematically illustrates the process of generating a vehicle maintenance scheduling strategy according to an embodiment of the present disclosure.

[0368] As shown in Figure 19, based on the ground maintenance operation status information 541D and the information of the parts to be maintained 542D, the waiting time 543D and the maintenance time 544D can be determined. Combined with the travel time 545D of the vehicle from its current location to the ground maintenance site, a maintenance scheduling strategy 546D is generated.

[0369] For example, based on the ground maintenance operation status information and the information of the parts to be maintained, the waiting time is determined to be 2 days, the maintenance time is 7 days, and the travel time for the vehicle from its current location to the ground maintenance site is 5 hours, thereby generating a maintenance scheduling strategy.

[0370] According to embodiments of this disclosure, the waiting time and maintenance time can be determined by using information on the components to be inspected and ground maintenance operation status information. By combining this with real-world application scenarios and comprehensively considering vehicle waiting time, maintenance time, and travel time, a maintenance scheduling strategy is generated. This makes the scheduling strategy more reasonable, efficient, and closer to actual conditions, facilitating smooth vehicle maintenance scheduling, preventing maintenance tracks from being occupied for extended periods, and improving maintenance efficiency.

[0371] According to embodiments of this disclosure, each waiting time and each maintenance time are determined based on information about each component to be inspected and information about the ground maintenance operation status. This includes: obtaining spare parts inventory information, historical procurement information, and historical maintenance information of the ground maintenance site; generating the procurement time for each component to be inspected based on the spare parts inventory information, information about each component to be inspected, and historical procurement information; obtaining the estimated maintenance time of vehicles waiting for inspection on each track from the ground maintenance operation status information; generating each waiting time based on the estimated maintenance time of vehicles waiting for inspection on each track and the procurement time for each component to be inspected; and generating each maintenance time based on information about each component to be inspected and historical maintenance information.

[0372] According to embodiments of this disclosure, the availability and procurement status of spare parts determine the time required for spare parts procurement, thus affecting the waiting time. Furthermore, the operational status of ground maintenance also influences the waiting time.

[0373] Figure 20 schematically illustrates the process of generating vehicle maintenance waiting time according to an embodiment of the present disclosure.

[0374] As shown in Figure 20, the waiting time 543D can be determined by the procurement time 543D-1 of the component to be repaired and the estimated repair time 543D-2. The procurement time 543D-1 of the component to be repaired is determined by the spare parts inventory information 543D-11, historical procurement information 543D-12, and the component information 54D2. The estimated repair time 543D-2 can be estimated using the ground maintenance operation status information 541D.

[0375] Spare parts inventory information 543D-11 includes the quantity and location of spare parts in stock. Historical procurement information 543D-12 includes the time required for the historical procurement process of the part to be repaired. Spare parts inventory information 543D-1 and historical procurement information 543D-2 can be obtained through communication between the various subsystems of ground maintenance. When the estimated waiting time 543D is used, if the part to be repaired needs to be replaced, it may be affected by the availability of spare parts for the part to be repaired.

[0376] For example: The component to be repaired is traction motor P1546, which needs to be replaced. The replacement quantity is 5. At this time, the spare parts inventory information for the same model of traction motor is "Traction motor P1546 in stock 20 pieces", which meets the replacement requirement. In this case, the material management personnel only need to retrieve the corresponding number of the required model of parts from the warehouse and deliver the materials. The estimated procurement time for the component to be repaired is 3 hours. If the spare parts inventory information for the same model of traction motor is "Traction motor P1546 in stock 0 pieces", which does not meet the replacement requirement, the component needs to be re-procured. In this case, based on the historical procurement information "Traction motor P1546 was ordered on October 5, 2023, and received into the warehouse on October 12, 2023, with a procurement time of 7 days", the estimated procurement time for the component to be repaired is 7 days.

[0377] Ground maintenance operation status information 541D includes personnel status, ground maintenance track occupancy status, and maintenance operation progress status information. Based on the ground maintenance operation status information, the estimated maintenance time 543D-2 can be determined. The estimated maintenance time represents the time taken for the vehicle to be maintained during ground maintenance operations.

[0378] For example: Currently, all three maintenance tracks on the ground are in operation, with progress rates of 5%, 60%, and 85% respectively, and maintenance personnel are all on duty. Based on the maintenance operation status analysis, the time required for vehicles on the three maintenance tracks to complete the current maintenance work is 7 days, 3 days, and 1 day respectively. Taking the minimum of the three, the estimated maintenance time is 1 day. If any of the maintenance tracks on the ground are currently idle, the estimated maintenance time is 0 days.

[0379] According to embodiments of this disclosure, based on the acquired spare parts inventory information, historical procurement information, and information on parts to be repaired, the procurement time for the parts to be repaired can be determined. Based on the ground maintenance operation status information, the estimated maintenance time can be determined. Based on the procurement time of the parts to be repaired and the estimated maintenance time, the waiting time for vehicle maintenance can be determined. By comprehensively considering the possible influencing factors of the waiting time, the estimation of the waiting time is made more accurate, and the generated maintenance scheduling strategy is more complete and reasonable, thereby improving maintenance efficiency.

[0380] According to embodiments of this disclosure, each maintenance duration is generated based on information about each component to be inspected and historical maintenance information, including: for each component to be inspected, extracting vehicle model information of the assembled component from the component information; determining at least one historical maintenance duration corresponding to the vehicle model information and the component information to be inspected from the historical maintenance information; and calculating the maintenance duration based on at least one historical maintenance duration.

[0381] According to embodiments of this disclosure, determining at least one historical maintenance duration corresponding to vehicle model information and component information to be maintained from historical maintenance information includes: historical maintenance durations corresponding to the same vehicle model information and the same component information to be maintained; historical maintenance durations corresponding to the same vehicle model information and the same type of component information to be maintained; historical maintenance durations corresponding to the same type of vehicle model information and the same component information to be maintained; and historical maintenance durations corresponding to the same type of vehicle model information and the same type of component information to be maintained. The historical maintenance information may contain historical maintenance information that completely matches the component information to be maintained and the vehicle model information, but it may also be impossible to find completely matching historical maintenance information. In this case, the maintenance duration can be determined based on the same type of component and model.

[0382] In particular, when multiple historical maintenance durations are obtained based on historical maintenance information, the maintenance duration can be calculated by weighted summation or by calculating the average of the historical maintenance durations.

[0383] According to embodiments of this disclosure, based on the information of the component to be repaired and the vehicle model information corresponding to the component, the corresponding historical repair duration is determined from historical repair information, and then the repair duration of the component to be repaired is estimated to generate a repair scheduling strategy. Since the repair scheduling strategy takes into account the actual repair situation, the resulting repair scheduling strategy is closer to the actual application scenario, improving the rationality of the repair scheduling strategy.

[0384] According to embodiments of this disclosure, the method further includes: determining each candidate track corresponding to the information of each component to be repaired; and determining the target track from each candidate track based on the operation status information on each candidate track.

[0385] Maintenance tracks include designated maintenance tracks for specific components, as well as general maintenance tracks. For example, maintenance track x is only used for sensor replacement, while maintenance track y can be used for major overhauls. It can be used for all maintenance processes, including vehicle disassembly and component disassembly, cleaning, assembly, testing, and vehicle final assembly and debugging. Therefore, it is necessary to determine the candidate tracks based on the condition of the component to be maintained.

[0386] Based on the operational status information of the candidate tracks, the target track is determined from among the candidate tracks to avoid excessively long maintenance waiting times and track occupancy conflicts. For example, if the operational status of candidate track A is idle, then that candidate track can be determined as the target track. If the operational status of candidate track A is active, then other candidate tracks can be determined as the target tracks.

[0387] According to embodiments of this disclosure, a corresponding candidate track is selected based on the condition of the component to be repaired, thereby achieving targeted and operable repairs. Determining the target track based on the operating status of the candidate tracks can reduce waiting time for repairs, avoid severe track occupancy, and improve repair efficiency.

[0388] According to embodiments of this disclosure, based on maintenance scheduling strategy information, scheduling each vehicle to enter a ground maintenance and repair site according to its corresponding track information to wait for maintenance includes: determining, based on the maintenance scheduling strategy information, the track information for each vehicle to enter its corresponding track and the time information for each vehicle to arrive at the ground maintenance site; sending the time information for each vehicle to arrive at the ground maintenance site and the track information for each vehicle to enter its corresponding track to each vehicle, so that each vehicle can generate a vehicle operation control strategy based on the time information, the track information and the vehicle's operating status information, so as to control each vehicle to enter the ground maintenance and repair site according to its corresponding track information to wait for maintenance.

[0389] According to embodiments of this disclosure, the arrival time information of each vehicle at the ground maintenance site can be determined based on the aforementioned travel time. After obtaining the arrival time information and the track information of the vehicle at the ground maintenance site, a vehicle operation scheduling strategy can be determined based on the aforementioned information, including when and where the vehicle departs, at what speed to travel the target distance, and when and where it arrives at the designated maintenance track.

[0390] According to embodiments of this disclosure, the ground maintenance system generates a corresponding vehicle maintenance scheduling strategy based on information obtained from the vehicle system and the ground maintenance subsystem, and sends the specific instructions of the vehicle maintenance scheduling strategy to the vehicle system to schedule vehicle operation.

[0391] The historical mileage and initial fault diagnosis information of each vehicle are analyzed to determine the information of each component to be repaired. The following describes the process of determining the component to be repaired, taking the fault occurring in the vehicle traction system as an example.

[0392] Figure 21 schematically illustrates an example judgment flowchart for determining the component to be repaired according to an embodiment of the present disclosure.

[0393] As shown in Figure 21, process 600D includes operations S621D to S625D.

[0394] When operating S621D, the historical mileage and initial fault diagnosis information of each vehicle are analyzed.

[0395] In operation S622D, it is determined whether the historical mileage of each vehicle is greater than a predetermined threshold. If the determination result is yes, operation S623 is executed; if the determination result is no, operation S624 is executed.

[0396] Using S623D, the component to be inspected was determined to be the traction motor bearing.

[0397] In operation S624D, determine whether the initial fault information is a traction motor grounding. If the determination result is yes, then execute operation S625.

[0398] When operating the S625D, it was determined that the component to be inspected was the traction motor.

[0399] For example, if vehicle X1235 has a historical mileage of 1.65 million kilometers, which is greater than the predetermined threshold of 1.5 million kilometers, then the component to be repaired can be identified as the traction motor bearing. If vehicle X1248 has a historical mileage of 1 million kilometers, which is less than the predetermined threshold of 1.5 million kilometers, and the initial fault information for this vehicle is traction motor grounding, then the component to be repaired can be identified as the traction motor.

[0400] According to embodiments of this disclosure, the condition of components to be repaired can be accurately and quickly determined by the vehicle's mileage and initial fault diagnosis information, so as to facilitate efficient vehicle maintenance in the future.

[0401] Based on the vehicle maintenance and scheduling method of the above-mentioned vehicle-to-ground integrated system, this disclosure also provides a vehicle maintenance and scheduling device for the vehicle-to-ground integrated system. The device will be described in detail below with reference to Figure 22.

[0402] Figure 22 schematically illustrates a structural block diagram of a vehicle maintenance and scheduling device based on a vehicle-to-ground integrated system according to an embodiment of the present disclosure.

[0403] As shown in Figure 22, the vehicle maintenance scheduling device 700D of this embodiment includes a first acquisition module 710D, an analysis module 720D, a calculation module 730D, a generation module 740D, and a scheduling module 750D.

[0404] The first acquisition module 710D is configured to acquire operating status information, initial fault diagnosis information, and ground maintenance operation status information of multiple vehicles. The operating status information includes operating location, operating speed, and historical mileage. In one embodiment, the first acquisition module 710D can be used to execute the operation S410D described above, which will not be repeated here.

[0405] The analysis module 720D is configured to analyze the historical mileage and initial fault diagnosis information of each vehicle to determine the information of each component to be repaired. In one embodiment, the analysis module 720D can be used to perform the operation S420D described above, which will not be repeated here.

[0406] The calculation module 730D is configured to calculate the travel time required for each vehicle to reach the ground maintenance and repair location based on the vehicle's operating location and speed. In one embodiment, the calculation module 730D can be used to perform the operation S430D described above, which will not be repeated here.

[0407] The generation module 740D is configured to generate maintenance scheduling strategy information for multiple vehicles based on information about each component to be inspected, each travel duration, and ground maintenance operation status information. The maintenance scheduling strategy information indicates the track information of each vehicle while waiting for maintenance at the ground maintenance site. In one embodiment, the generation module 740D can be used to execute the operation S440D described above, which will not be repeated here.

[0408] The scheduling module 750D is configured to schedule each vehicle to enter the ground maintenance area to await maintenance according to its corresponding track information based on maintenance scheduling strategy information. In one embodiment, the scheduling module 750D can be used to execute the operation S450D described above, which will not be repeated here.

[0409] According to embodiments of this disclosure, the generation module 740D includes a duration determination submodule and a generation submodule. The duration determination submodule is configured to determine each waiting duration and each maintenance duration based on information about each component to be maintained and ground maintenance operation status information; wherein each waiting duration represents the time required for each vehicle to wait for maintenance; and each maintenance duration represents the time required to maintain each vehicle. The generation submodule is configured to generate maintenance scheduling strategy information based on each waiting duration, each maintenance duration, and each travel duration.

[0410] According to embodiments of this disclosure, the duration determination submodule includes an information acquisition unit, a procurement duration generation unit, an estimated maintenance duration acquisition unit, a waiting duration generation unit, and a maintenance duration generation unit. The information acquisition unit is configured to acquire spare parts inventory information, historical procurement information, and historical maintenance information of the ground maintenance site. The procurement duration generation unit is configured to generate the procurement duration for each component to be maintained based on the spare parts inventory information, information on each component to be maintained, and historical procurement information. The estimated maintenance duration acquisition unit is configured to acquire the estimated maintenance duration of vehicles waiting for inspection on each track from the ground maintenance operation status information. The waiting duration generation unit is configured to generate each waiting duration based on the estimated maintenance duration of vehicles waiting for inspection on each track and the procurement duration of each component to be maintained. The maintenance duration generation unit is configured to generate each maintenance duration based on information on each component to be maintained and historical maintenance information.

[0411] According to embodiments of this disclosure, the maintenance time generation unit includes a model extraction subunit, a historical maintenance time determination subunit, and a maintenance time calculation subunit. The model extraction subunit is configured to extract, for each component to be maintained, the vehicle model information of the assembled component from the component information. The historical maintenance time determination subunit is configured to determine at least one historical maintenance time corresponding to the vehicle model information and the component information to be maintained from historical maintenance information. The maintenance time calculation subunit is configured to calculate the maintenance time based on at least one historical maintenance time.

[0412] According to embodiments of this disclosure, the above-described apparatus further includes a candidate track determination module and a target track determination module. The candidate track determination module is configured to determine each candidate track corresponding to the information of each component to be repaired. The target track determination module is configured to determine a target track from the candidate tracks based on the operation status information on each candidate track.

[0413] According to embodiments of this disclosure, the scheduling module 750D includes an information determination submodule and a sending submodule. The information determination submodule is configured to determine, based on maintenance scheduling strategy information, the entry information of each vehicle into its corresponding track and the arrival time information of each vehicle at the ground maintenance site. The sending submodule is configured to send the arrival time information of each vehicle at the ground maintenance site and the entry information of each vehicle into its corresponding track to each vehicle, so that each vehicle can generate a vehicle operation control strategy based on the time information, track information, and vehicle operating status information, thereby controlling each vehicle to enter the ground maintenance site according to its corresponding track information to wait for maintenance.

[0414] According to embodiments of this disclosure, the analysis module 720D includes a first response submodule and a second response submodule. The first response submodule is configured to determine that the component to be repaired is a traction motor bearing in response to a historical operating mileage exceeding a predetermined threshold. The second response submodule is configured to determine that the component to be repaired is a traction motor in response to a historical operating mileage less than or equal to a predetermined threshold and an initial fault diagnosis message indicating traction motor grounding.

[0415] Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure, or at least part of the functions of any one or more of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be implemented by dividing them into multiple modules. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as hardware circuitry, such as Field Programmable Gate Arrays (FPGAs), Programmable Logic Arrays (PLAs), Systems-on-Chip, Systems-on-Substrate, Systems-on-Package, Application-Specific Integrated Circuits (ASICs), or implemented in hardware or firmware by any other reasonable means of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0416] For example, any and more of the first acquisition module 710D, analysis module 720D, calculation module 730D, generation module 740D, and scheduling module 750D can be combined into one module / unit / subunit, or any one of these modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least some of the functions of one or more of these modules / units / subunits can be combined with at least some of the functions of other modules / units / subunits and implemented in one module / unit / subunit. According to embodiments of this disclosure, at least one of the first acquisition module 710D, analysis module 720D, calculation module 730D, generation module 740D, and scheduling module 750D can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or any other reasonable method of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three methods. Alternatively, at least one of the first acquisition module 710D, analysis module 720D, calculation module 730D, generation module 740D, and scheduling module 750D can be at least partially implemented as a computer program module, which, when run, can perform corresponding functions.

[0417] Based on the vehicle maintenance and scheduling method of the above-mentioned vehicle-ground integrated system, this disclosure also provides a train, including a second acquisition module and a control module. The above modules will be described in detail below with reference to Figure 23.

[0418] Figure 23 schematically illustrates a structural block diagram of a train module according to an embodiment of the present disclosure.

[0419] As shown in Figure 23, the train 800D in this embodiment includes a second acquisition module 810D and a control module 820D.

[0420] The second acquisition module 810D is configured to acquire scheduling instructions. These instructions are generated based on the vehicle maintenance scheduling method of the aforementioned vehicle-ground integrated system, which produces maintenance scheduling strategy information. The control module 820D is configured to generate a vehicle control strategy based on the aforementioned scheduling instructions, controlling the train to enter the maintenance area according to the track information indicated in the maintenance scheduling strategy information to wait for maintenance. The vehicle control strategy can be implemented through a strategy generation algorithm, which can be determined based on the actual application scenario; no restrictions are placed on the strategy generation algorithm here.

[0421] Figure 24 schematically illustrates a block diagram of an electronic device suitable for implementing any of the methods described above, according to embodiments of the present disclosure.

[0422] As shown in FIG24, an electronic device 2400 according to an embodiment of the present disclosure includes a processor 2401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 2402 or a program loaded from a storage portion 2408 into a random access memory (RAM) 2403. The processor 2401 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 2401 may also include onboard memory for caching purposes. The processor 2401 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0423] RAM 2403 stores various programs and data required for the operation of electronic device 2400. Processor 2401, ROM 2402, and RAM 2403 are interconnected via bus 2404. Processor 2401 performs various operations of the method flow according to embodiments of the present disclosure by executing programs in ROM 2402 and / or RAM 2403. It should be noted that programs may also be stored in one or more memories other than ROM 2402 and RAM 2403. Processor 2401 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in one or more memories.

[0424] According to embodiments of this disclosure, the electronic device 2400 may further include an input / output (I / O) interface 2405, which is also connected to a bus 2404. The system 2400 may also include one or more of the following components connected to the input / output (I / O) interface 2405: an input section 2406 including a keyboard, mouse, etc.; an output section 2409 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 2408 including a hard disk, etc.; and a communication section 2409 including a network interface card such as a LAN card, modem, etc. The communication section 2409 performs communication processing via a network such as the Internet. A drive 2410 is also connected to the input / output (I / O) interface 2405 as needed. A removable medium 2411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 2410 as needed so that computer programs read from it can be installed into the storage section 2408 as needed.

[0425] According to embodiments of this disclosure, the method flow according to embodiments of this disclosure can be implemented as a computer software program. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 2409, and / or installed from removable medium 2411. When the computer program is executed by processor 2401, it performs the functions defined in the system of embodiments of this disclosure. According to embodiments of this disclosure, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0426] This disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs that, when executed, implement the method according to the embodiments of this disclosure.

[0427] According to embodiments of this disclosure, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0428] For example, according to embodiments of this disclosure, a computer-readable storage medium may include one or more memories other than the ROM 2402 and / or RAM 2403 described above.

[0429] Embodiments of this disclosure also include a computer program product comprising a computer program containing program code for performing the methods provided in the embodiments of this disclosure. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the methods provided in the embodiments of this disclosure.

[0430] When the computer program is executed by the processor 2401, it performs the functions defined in the system / apparatus of this disclosure embodiments. According to embodiments of this disclosure, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0431] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 2409, and / or installed from the removable medium 2411. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0432] According to embodiments of this disclosure, program code for executing the computer programs provided in embodiments of this disclosure can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can execute entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0433] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features recited in the various embodiments and / or claims of this disclosure can be combined and / or combined in various ways, even if such combinations or combinations are not expressly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this disclosure may be combined and / or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0434] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this disclosure is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A fault determination method based on a track-tunnel system, comprising: The first real-time operating status information of the target train during the target time period and the first real-time operating environment information of the line it travels on are obtained. The first real-time operating environment information includes real-time pantograph-catenary monitoring data, real-time track monitoring data and real-time tunnel monitoring data collected by the train monitoring equipment at the same time and space as the real-time operating status information. Based on a predetermined map, the target correlation between the real-time operating status information, the real-time pantograph-catenary monitoring data, the real-time track monitoring data, and the real-time tunnel monitoring data is determined. The predetermined map is constructed based on historical pantograph-catenary data, historical track data, and historical tunnel data, and is used to characterize the correlation between the historical pantograph-catenary data, the historical track data, and the historical tunnel data. The target association is input into the trained target graph neural network model, which outputs potential fault information related to the operating environment of the target train. The potential fault information includes at least potential fault points and fault occurrence probabilities, so that fault troubleshooting operations can be performed on the potential fault points before the target train reaches them.

2. The method according to claim 1, wherein, The predetermined map is constructed using the following method: The historical operating status information and historical operating environment information of the target train within a historical time period are obtained. The historical operating status information includes historical driving stability data, and the real-time operating environment information includes historical pantograph-catenary data, historical track data, and historical tunnel data collected by the train monitoring equipment at the same time and space as the historical operating status information. Based on the L types of data contained in the historical driving stability data, the historical pantograph-catenary data, the historical track data, and the historical tunnel data, at least one transaction dataset is constructed, wherein the at least one transaction dataset contains L transactions corresponding to the L types of data; Based on the support and support threshold of each transaction, multiple frequent itemsets are determined; Based on each transaction in each frequent itemset, generate multiple frequent sub-itemsets corresponding to the frequent itemsets; Calculate the lift of each frequent subset and determine the association between transactions in the frequent subset based on the lift. Based on the relationships between the various transactions, a predetermined graph of each transaction is constructed.

3. The method according to claim 1, wherein, The L data types include: The historical driving stability data includes at least one or a combination of the following data types: the target train's smoothness detection data, instability detection data, vibration detection data, and idle coasting data; The historical pantograph-catenary data includes at least one or a combination of the following data types: catenary pressure data, catenary current data, pantograph-catenary contact pressure data, track arcing rate data, and pantograph-catenary foreign object data of the target train. The historical track data includes at least one or a combination of the following data types: rail profile data, rail corrugation data, turnout abnormal status data, track gauge detection data, and rail surface status data during idle sliding. The historical tunnel data includes at least one or a combination of the following data types: tunnel foreign object intrusion data, tunnel clearance data, and tunnel video anomaly data.

4. The method according to claim 1, wherein, The target graph neural network model is trained using the following method: Obtain the predetermined map and historical fault diagnosis information within the historical time period; Using the predetermined graph as historical sample data and the historical fault diagnosis information as labels, the initial graph neural network model is optimized to obtain the target graph neural network model.

5. The method according to claim 1, wherein, The method further includes: Based on the second real-time operating status information of multiple reference trains in the same area, the second real-time operating environment information, and the probability of the fault occurrence, a target fault point is determined from the potential fault points so that the target fault point can be eliminated in advance.

6. The method according to claim 1, wherein, The potential fault information also includes the potential fault type and the expected occurrence time of the potential fault, and the method further includes: Based on the potential failure point, the potential failure type, and the expected occurrence time of the potential failure, determine the severity and scope of impact of the potential failure; Based on the severity and scope of impact, determine the processing priority and processing strategy for the potential fault.

7. A train, comprising: Body; A braking component and a fault determination device based on a track-tunnel system, wherein the fault determination device based on the track-tunnel system includes: The acquisition module is configured to acquire the first real-time operating status information of the target train during the target time period and the first real-time operating environment information of the line it travels on. The first real-time operating environment information includes real-time pantograph-catenary monitoring data, real-time track monitoring data and real-time tunnel monitoring data collected by the train monitoring equipment at the same time and space as the real-time operating status information. The determination module is configured to determine the target correlation between the real-time operating status information, the real-time pantograph-catenary monitoring data, the real-time track monitoring data, and the real-time tunnel monitoring data based on a predetermined map. The predetermined map is constructed based on historical pantograph-catenary data, historical track data, and historical tunnel data, and is used to characterize the correlation between the historical pantograph-catenary data, the historical track data, and the historical tunnel data. The output module is configured to input the target association into a trained target graph neural network model and output potential fault information related to the operating environment of the target train. The potential fault information includes at least potential fault points and fault occurrence probabilities, so that fault troubleshooting operations can be performed on the potential fault points before the target train reaches them.

8. A train operation scheduling method based on a network track tunnel system, comprising: Acquire potential fault information detected when the reference train is running on the initial line, wherein the potential fault information includes one or more of the following: catenary fault information, track fault information, and tunnel fault information related to the operating environment of the initial line; The first operating status information, first operating mileage information, first maintenance information, and multiple first lines to be operated information of the target train are obtained within a first historical time period. The target train is a train that is planned to travel on the initial line within a target preset time period. Based on the first operating status information, the first operating mileage information, and the first maintenance information, the first health status information of the target train is generated; Based on the first health status information and the potential fault information, a first operation strategy is generated for the target train to control the target train to continue running on the initial line or to dispatch the target train to a first alternative line other than the initial line from the plurality of first lines to be run information.

9. The method according to claim 8, wherein, Based on the first operating status information, the first operating mileage information, and the first maintenance information, the first health status information of the target train is generated, including: Based on the first operating status information, the first operating mileage information, and the first maintenance information, the target train's energy consumption information and loss information are generated. Based on the energy consumption information and the loss information, the first health status information of the target train is generated.

10. The method according to claim 9, wherein, Based on the first operating status information, the first operating mileage information, and the first maintenance information, the energy consumption information and loss information of the target train are generated, including: Input the first operating status information, the first operating mileage information, and the first maintenance information into the trained first target model, and output the energy consumption information and the loss information; The first target model is obtained by training the first initial model using the historical energy consumption information and historical loss information of the sample train as labels, and using the historical operating status information, historical mileage information and historical maintenance information of the sample train.

11. The method according to claim 9, wherein, Based on the energy consumption information and the loss information, the first health status information of the target train is generated, including: The energy consumption information, the loss information, and the first operating mileage information are input into the trained second target model, and the health status information of the target train is output. The second target model is obtained by training the second initial model using the historical failure rate of each sample train as a label, and the historical energy consumption information, the historical loss information, and the historical running mileage information.

12. The method according to claim 8, wherein, The potential fault information includes potential fault points, and the health status information includes healthy and unhealthy states. Based on the first health status information and the potential fault information, a first operating strategy is generated for the target train to control the target train to continue operating on the initial line or to dispatch the target train to a first alternative line other than the initial line from the plurality of first lines to be operated information, including: The processing level of the potential fault points is determined based on the potential fault information, wherein the processing level includes urgent and non-urgent. If the first health status information of the target train is healthy and the pending level is non-emergency, the target train shall be controlled to continue running on the initial line. If the first health status information of the target train is healthy and the pending level is emergency, the target train will be dispatched to one of the first alternative lines to run. If the first health status information of the target train is unhealthy, the target train will be dispatched to one of the first alternative lines.

13. The method according to claim 8 or 12, wherein, When the first operating strategy is to dispatch the target train to the first alternative line, the method further includes: Obtain the operating environment monitoring information of each of the first candidate lines; Based on the first health status information of the target train and the operating environment monitoring information of each of the first alternative routes, a scheduling plan is generated for the target train to be scheduled to run on one of the first alternative routes.

14. The method according to claim 8, wherein, The method further includes: Obtain the second operating status information, second operating mileage information, second maintenance information, and multiple second lines to be operated information of the reference train within the second historical time period; Based on the second operating status information, the second operating mileage information, and the second maintenance information, the second health status information of the reference train is generated; Based on the second health status information and the potential fault information, a second operating strategy is generated for the reference train to control the reference train to continue running on the initial line or to dispatch the reference train to a second alternative line other than the initial line from the plurality of second alternative lines to be run.

15. A train comprising: The acquisition module is configured to acquire the first operating strategy and / or the second operating strategy as described in any one of claims 8 to 14; The generation module is configured to generate operation instructions based on the first operation strategy and / or the second operation strategy to control the train to operate in accordance with the first operation strategy and / or the second operation strategy.

16. A train operation scheduling method based on a vehicle-ground integrated system, comprising: Acquire status monitoring information of each component configured on multiple candidate trains, mileage information of each component, maintenance record information of each component, historical operating environment information of multiple candidate trains, and monitoring information of multiple lines to be operated; Based on the status monitoring information of each component, the operating mileage information of each component, the maintenance record information of each component, and the historical operating environment information of each candidate train, health status information of each candidate train is generated. Based on the monitoring information of each of the lines to be operated, generate the status information of each of the lines to be operated. as well as Based on the health status information of each candidate train and the status information of each line to be operated, an operation scheduling plan is generated for each candidate train to schedule each candidate train to run on each line to be operated.

17. The method according to claim 16, wherein, The health status information of each candidate train is generated based on the status monitoring information of each component, the mileage information of each component, the maintenance record information of each component, and the historical operating environment information of each candidate train, including: Extract the historical operating environment information of each component from the historical operating environment information of each candidate train; Based on the mileage information, historical operating environment information, and maintenance record information of each component, the remaining lifespan information of each component is generated; and Based on the remaining lifespan information of each component of the same candidate train and the status monitoring information of each component, health status information of each candidate train is generated.

18. The method according to claim 17, wherein, Based on the mileage information, historical operating environment information, and maintenance record information of each component, the remaining lifespan information of each component is generated, including: The mileage information of each component, the historical operating environment information of each component, and the maintenance record information of each component are input into the trained first target model, and the remaining life information of each component is output respectively. The first target model is obtained by training the first initial model using the historical remaining life of the sample component as a label and utilizing historical operating mileage, historical operating environment and historical maintenance records.

19. The method of claim 17, wherein, The process of generating health status information for each candidate train based on the remaining lifespan information and the status monitoring information of each component of the same candidate train includes: The remaining life information and the status monitoring information of each component of the same candidate train are input into the trained second target model, and the health status information of each candidate train is output respectively. The second target model is obtained by training the second initial model using the historical failure rate of each sample train as a label and the historical remaining lifespan and historical monitoring status of the components of each sample train.

20. The method of claim 16, wherein, The monitoring information for the line to be put into operation includes the monitoring information of the overhead contact system, track monitoring information, and tunnel monitoring information corresponding to the line to be put into operation; The step of generating status information for each of the lines to be operated based on the monitoring information of each line includes: The monitoring information of each overhead contact line, each track, and each tunnel corresponding to each of the lines to be operated are input into the trained third target model to generate the status information of each of the lines to be operated. The third target model is obtained by training the third initial model using the status of the sample running line as a label and the historical monitoring information of the overhead contact line, the track, and the tunnel corresponding to the sample running line.

21. The method according to claim 16, wherein, The step of generating an operation scheduling plan for each candidate train based on the health status information of each candidate train and the status information of each line to be operated, so as to schedule each candidate train to run on each line to be operated, includes: The initial matching result is obtained by matching the health status information of each candidate train with the status information of each line to be operated. The initial matching results are verified based on the remaining lifespan information of each component in each of the candidate trains to obtain verification results; and In response to the verification result being passed, the running schedule plan is generated based on the initial matching result.

22. A train, comprising: The acquisition module is configured to acquire the operation scheduling plan as described in any one of claims 16 to 21; The generation module is configured to generate operation instructions based on the operation scheduling plan in order to control the train to operate in accordance with the operation scheduling plan.

23. A vehicle maintenance scheduling method based on a vehicle-to-ground integrated system, comprising: The system acquires operational status information, initial fault diagnosis information, and ground maintenance operation status information for multiple vehicles; wherein the operational status information includes operating location, operating speed, and historical operating mileage. The historical mileage and initial fault diagnosis information of each vehicle are analyzed to determine the information of each component to be repaired. Based on the stated operating location and speed of each vehicle, calculate the travel time required for each vehicle to reach the ground inspection and maintenance site. Based on the information of each component to be inspected, the travel time, and the ground maintenance operation status, maintenance scheduling strategy information is generated for the multiple vehicles; wherein, the maintenance scheduling strategy information indicates the track information of each vehicle while waiting for maintenance at the ground maintenance site; and Based on the maintenance scheduling strategy information, each vehicle is dispatched to enter the ground maintenance site according to its corresponding track information to wait for maintenance.

24. The method according to claim 23, wherein, The step of generating maintenance scheduling strategy information for the multiple vehicles based on the information of each component to be inspected, the driving time of each component, and the ground maintenance operation status information includes: Based on the information of each component to be inspected and the status information of the ground maintenance operation, each waiting time and each maintenance time are determined; wherein, each waiting time represents the time required for each vehicle to wait for maintenance; and each maintenance time represents the time required for each vehicle to be maintained. The maintenance scheduling strategy information is generated based on the waiting time, maintenance time, and travel time.

25. The method according to claim 24, wherein, The step of determining each waiting time and each maintenance time based on the information of each component to be inspected and the ground maintenance operation status information includes: Obtain spare parts inventory information, historical procurement information, and historical maintenance information of the ground maintenance and repair sites; Based on the spare parts inventory information, the information of each component to be repaired, and historical procurement information, the procurement duration of each component to be repaired is generated; The estimated maintenance time of vehicles waiting for inspection on each track is obtained from the ground maintenance operation status information. Based on the estimated maintenance time of vehicles waiting for inspection on each track and the procurement time of each component to be inspected, the waiting time for each item is generated; and Based on the information of each component to be inspected and the historical inspection information, the inspection duration is generated.

26. The method of claim 25, wherein, The step of generating each maintenance duration based on the information of each component to be inspected and the historical maintenance information includes: For each component to be repaired, extract the vehicle model information of the component already assembled with it from the component information. Determine at least one historical maintenance duration from the historical maintenance information that corresponds to the vehicle model information and the information of the component to be maintained; and The maintenance duration is calculated based on the at least one historical maintenance duration.

27. The method according to claim 24, wherein, The method further includes: Determine each candidate track corresponding to the information of each component to be inspected; Based on the operational status information of each candidate track, the target track is determined from among the candidate tracks.

28. The method according to claim 24, wherein, The step of scheduling each vehicle to enter the ground maintenance and repair site according to its corresponding track information based on the maintenance and repair strategy information includes: Based on the maintenance scheduling strategy information, the information on each vehicle entering its corresponding track and the time information of each vehicle arriving at the ground maintenance site are determined. The system sends the arrival time information of each vehicle at the ground maintenance site and the entry information of each vehicle into its corresponding lane to each vehicle. This allows each vehicle to generate a vehicle operation control strategy based on the arrival time information, the lane information, and the vehicle's operating status information, so as to control each vehicle to enter the ground maintenance site according to its corresponding lane information to wait for maintenance.

29. The method according to claim 23, wherein, The analysis of the historical mileage and initial fault diagnosis information of each vehicle to determine the information of each component to be repaired includes: In response to the historical operating mileage exceeding a predetermined threshold, the component to be inspected is determined to be a traction motor bearing; and In response to the historical operating mileage being less than or equal to a predetermined threshold and the initial fault diagnosis information indicating that the traction motor is grounded, the component to be repaired is determined to be the traction motor.

30. A train comprising: The acquisition module is configured to acquire scheduling instructions; wherein the scheduling instructions are generated based on maintenance scheduling strategy information generated by the method described in any one of claims 23 to 29. The control module is configured to generate a vehicle control strategy based on the scheduling instructions, so as to control the train to enter the ground maintenance site for maintenance according to the track information indicated in the maintenance scheduling strategy information.

Images