Rail traffic vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph

Abstract

The invention belongs to the technical field of rail transit vehicles, and discloses a rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on a knowledge graph, and the service recommendation system is composed of a rail transit vehicle fault detection module, a knowledge graph construction and updating module, and a maintenance service recommendation module. According to the method, efficient, accurate and real-time monitoring and fault diagnosis of the vehicle state are realized through a fault detection algorithm, powerful data support is provided for intelligent fault diagnosis through a constructed comprehensive, dynamic and highly integrated rail transit vehicle knowledge graph, a reliable knowledge basis is provided for maintenance service recommendation, and the reliability of fault diagnosis is improved. The method greatly improves the safety and maintenance efficiency of the rail transit system, accurately generates and recommends an optimal maintenance service scheme through a maintenance service recommendation algorithm, achieves the intelligent recommendation of the maintenance service, improves the maintenance efficiency to the maximum degree, and reduces the operation cost.

Description

Technical Field

[0001] The present invention belongs to the technical field of rail transit vehicles, and specifically is a rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph. Background Art

[0002] Rail transit is a means of transportation or a transportation system that uses rail trains to transport people. It covers many types, including but not limited to subways, light rails, trams, maglev trains, and traditional railway systems. These systems are usually driven by electricity or other energy sources and run along specific laid tracks to achieve fast, efficient, and large-capacity transportation services.

[0003] At present, rail transit plays a pivotal role in modern society. It not only improves traffic conditions, promotes economic development, and enhances the image of cities, but also makes important contributions to environmental protection and sustainable development. However, with the long-term operation of rail transit vehicles, vehicle failures and vehicle maintenance issues have become increasingly prominent.

[0004] In rail transit operation management, vehicle fault diagnosis and maintenance services are key links to ensure the safety and stable operation of the system. However, traditional maintenance methods rely too much on manual inspections and expert experience, which not only leads to low diagnostic efficiency, but also a high risk of missed detection, posing a potential threat to operational safety and affecting the normal operation of rail transit. Therefore, a rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system is needed to assist maintenance personnel in diagnosis. Summary of the invention

[0005] The purpose of the present invention is to provide a rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph to solve the problems raised in the above background technology.

[0006] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: a rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph, wherein the service recommendation system is composed of a rail transit vehicle fault detection module, a knowledge graph construction and update module, and a maintenance service recommendation module;

[0007] The rail transit vehicle fault detection module is implemented based on a fault detection algorithm, the knowledge graph construction and update module includes the construction of a knowledge graph and the real-time update of a knowledge graph, and the maintenance service recommendation module is implemented based on a maintenance service recommendation algorithm.

[0008] Preferably, the fault detection algorithm includes data collection and preprocessing, feature extraction, deep learning model training, continuous learning and updating, real-time monitoring and fault identification and fault diagnosis result output, and the specific process of the fault detection algorithm is:

[0009] A1, first collect and organize the historical data of rail transit vehicles, and standardize the data;

[0010] A2, through the efficient processing of huge sensor data sets through deep learning technology, and then accurately extracting significant features;

[0011] A3, then, make full use of historical sensor data to train the deep learning model, and finally build an accurate vehicle status monitoring and potential fault prediction model through repeated iteration and optimization;

[0012] A4, then introduces an adaptive continuous learning and updating mechanism. During the actual operation of the vehicle, the deep learning model is incrementally learned online using real-time streaming monitoring data to dynamically optimize the model parameters.

[0013] A5, at this point the model has the ability to continuously and in real time monitor the vehicle sensor data stream, and based on the real-time input data, the built-in algorithm can efficiently identify abnormal patterns and accurately indicate the potential source of failure;

[0014] A6,Finally, the algorithm will output the accurately identified failure mode and its related information in a structured form for subsequent advanced processing and dedicated maintenance service recommendation algorithms.

[0015] Preferably, the specific process of data collection and preprocessing in the fault detection algorithm is:

[0016] B1, firstly, adopt data integration technology to systematically collect and organize the historical data of rail transit vehicles;

[0017] B2, then preprocess the data, and in the data preprocessing stage, use data cleaning technology to eliminate redundancy, errors and outliers in the data to ensure the accuracy and reliability of the data;

[0018] B3, then use the denoising algorithm to reduce the noise interference in the sensor data and improve the signal-to-noise ratio of the data;

[0019] B4. Finally, through data standardization, we ensure that data from different sources and scales can be compared and analyzed under the same framework, thereby ensuring the quality and consistency of the calculation data.

[0020] Preferably, the method for constructing the knowledge graph is:

[0021] C1, first, extensively collect relevant data on rail transit vehicles to ensure the comprehensiveness and diversity of the data. These data can come from real-time sensor data streams, organized maintenance record databases, and technical documents and parameter tables obtained directly from manufacturers;

[0022] C2, after obtaining the original data, execute the data annotation process, accurately classify each data according to its characteristics and relevance through data annotation technology, and build the corresponding data model to provide a structured data foundation for the subsequent knowledge graph construction and ensure the logic and consistency of the data;

[0023] C3, finally, the knowledge graph architecture is designed through the constructed data model, and the hierarchical structure of rail transit vehicles and the correlation between internal components are fully considered during the design to ensure that the knowledge graph can fully and accurately reflect the complexity and dynamics of the vehicle system;

[0024] C4, finally, accurately define the entities, relationships, and attributes in the graph to complete the graph model design.

[0025] Preferably, the real-time updating method of the knowledge graph is:

[0026] D1, first build a refined data interface and automatic crawling mechanism to collect real-time data efficiently and automatically;

[0027] D2, cleans, removes duplications, and supplements missing information of the captured real-time data, so that the data can be efficiently integrated with the existing data in the knowledge graph;

[0028] D3, through database storage technology, accurately associates the processed and updated data with the knowledge graph, thereby achieving real-time updating of the knowledge graph.

[0029] Preferably, in the construction and update module of the knowledge graph, when constructing the knowledge graph, it is necessary to associate the complex hierarchical structure of the rail transit vehicle, including the hierarchical relationship from the vehicle as a whole to each component, and by defining the interaction and dependency relationship between them, the knowledge graph is made more instructive and practical in the actual application of rail transit vehicles;

[0030] In the construction and update module of the knowledge graph, an efficient update mechanism is used to ensure that the knowledge graph can quickly capture and integrate changes in new data during updating, thereby improving its dynamic response capability.

[0031] Preferably, the specific process of the maintenance service recommendation algorithm is:

[0032] E1, when the vehicle system detects an abnormal condition, the system fault detection mechanism will be triggered immediately, thus starting a detailed fault detection process;

[0033] E2, accurately maps the detected fault information into the constructed knowledge graph structure, and then deeply analyzes the potential fault mode and possible impact range;

[0034] E3, at this time, the maintenance service recommendation algorithm directly accesses the vehicle's maintenance record database through an efficient data interface to obtain detailed historical maintenance records of the vehicle, and accurately identifies the processing mode and experience of similar faults in the past, thereby providing a more accurate and effective recommendation plan for the current fault repair;

[0035] E4, at the same time, the algorithm will obtain comprehensive information of the maintenance team in real time for accurate capture and integration, so that it can provide more accurate and efficient personnel allocation suggestions for fault maintenance tasks based on the actual capabilities and current status of the maintenance personnel;

[0036] E5, the final algorithm performs a series of evaluation operations to accurately determine the feasibility of different maintenance solutions by comprehensively considering the fault details, the in-depth analysis results of the knowledge graph, the historical maintenance records, and the real-time maintenance team status;

[0037] E6, at the same time, through advanced optimization algorithms, potential maintenance solutions are arranged, combined and iteratively optimized to find the global optimal solution;

[0038] E7,Finally, the algorithm will generate a detailed and optimized maintenance service recommendation plan, which includes a series of specific maintenance operation steps, estimated completion time, and detailed information of professional maintenance personnel assigned for each task.

[0039] Preferably, the service recommendation system collects detailed feedback information in real time through a data capture mechanism during the maintenance process. The collected high-quality data will be used to dynamically adjust the algorithm, thereby continuously optimizing the recommendation model and improving the system's adaptability and prediction accuracy.

[0040] The beneficial effects of the present invention are as follows:

[0041] 1. The present invention realizes efficient, accurate and real-time monitoring and fault diagnosis of vehicle status through fault detection algorithm, which provides strong technical support for improving the operational safety of rail transit vehicles, reducing maintenance costs and optimizing maintenance services. The comprehensive, dynamic and highly integrated rail transit vehicle knowledge graph constructed provides strong data support for intelligent fault diagnosis, and also provides a reliable knowledge basis for maintenance service recommendation, which greatly improves the safety and maintenance efficiency of the rail transit system. Finally, the maintenance service recommendation algorithm accurately generates and recommends the most optimized maintenance service plan, realizes intelligent recommendation of maintenance services, and maximizes maintenance efficiency while reducing operating costs.

[0042] 2. In the process of improving the rail transit vehicle fault diagnosis and maintenance service recommendation method, the present invention adopts advanced machine learning algorithms, especially deep learning and pattern recognition technology, to achieve real-time monitoring of vehicle status and accurate identification of potential faults. These cutting-edge algorithms are trained on massive historical data and have the ability to adapt to various complex vehicle operating conditions in continuous learning and updating.

[0043] 3. The present invention supports more accurate fault diagnosis by constructing a rail transit vehicle knowledge graph that covers key data such as detailed information of vehicle components, fault mode classification, and maintenance history records. At the same time, by using data capture and processing technology, the knowledge graph is updated in real time to ensure that it remains synchronized with the vehicle status, thereby reflecting the latest status of vehicle operation and ensuring timeliness. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The present invention is a service recommendation system operation flow chart. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] like Figure 1 As shown, an embodiment of the present invention provides a rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on a knowledge graph, wherein the service recommendation system is composed of a rail transit vehicle fault detection module, a knowledge graph construction and update module, and a maintenance service recommendation module;

[0047] The rail transit vehicle fault detection module is implemented based on a fault detection algorithm, the knowledge graph construction and update module includes the construction of a knowledge graph and the real-time update of a knowledge graph, and the maintenance service recommendation module is implemented based on a maintenance service recommendation algorithm.

[0048] Through the fault detection algorithm, efficient, accurate and real-time monitoring and fault diagnosis of vehicle status are achieved, which provides strong technical support for improving the operating safety of rail transit vehicles, reducing maintenance costs and optimizing maintenance services. The comprehensive, dynamic and highly integrated rail transit vehicle knowledge graph constructed provides strong data support for intelligent fault diagnosis and a reliable knowledge basis for maintenance service recommendations, which greatly improves the safety and maintenance efficiency of the rail transit system. Finally, the maintenance service recommendation algorithm accurately generates and recommends the most optimized maintenance service plan, realizing intelligent recommendation of maintenance services.

[0049] The fault detection algorithm includes data collection and preprocessing, feature extraction, deep learning model training, continuous learning and updating, real-time monitoring and fault identification and fault diagnosis result output, and the specific process of the fault detection algorithm is as follows:

[0050] A1, first collect and organize the historical data of rail transit vehicles, and standardize the data;

[0051] A2, through the efficient processing of huge sensor data sets through deep learning technology, and then accurately extracting significant features;

[0052] These features cover the multi-state information of the vehicle, including but not limited to the vibration spectrum characteristics of the vehicle, real-time temperature data, voltage fluctuations, and current changes. The core goal of feature extraction is to accurately capture the key features that are closely related to vehicle state changes and potential failure modes, thereby providing strong data support for subsequent vehicle state monitoring, performance evaluation, and fault diagnosis.

[0053] A3, then, make full use of historical sensor data to train deep learning models, including deep neural networks (DNN), convolutional neural networks (CNN) and recurrent neural networks (RNN), and finally build accurate vehicle status monitoring and potential fault prediction models through repeated iterations and optimization;

[0054] These deep learning models automatically learn the inherent patterns and potential laws in the data through deep learning algorithms, and the vehicle condition monitoring and potential fault prediction models ultimately constructed can accurately capture the changing trends of vehicle conditions and predict possible failures, providing strong technical support for vehicle maintenance and fault prevention.

[0055] A4, then introduces an adaptive continuous learning and updating mechanism. During the actual operation of the vehicle, the deep learning model is incrementally learned online using real-time streaming monitoring data to dynamically optimize the model parameters.

[0056] This mechanism ensures that the model can quickly adapt to new data distributions, thereby more accurately identifying emerging failure modes. This continuous learning and updating strategy not only enhances the timeliness of the algorithm, but also significantly improves its robustness in a changing environment.

[0057] A5, at this point, the model has gone through a careful training and deployment process and has the ability to continuously and in real time monitor the vehicle sensor data stream. Based on the input data acquired in real time, the built-in algorithm can efficiently identify abnormal patterns and accurately indicate potential sources of failure;

[0058] This critical information is then used in a timely maintenance warning mechanism, significantly reducing the risk of vehicle failures.

[0059] A6,Finally, the algorithm will output the accurately identified failure mode and its related information in a structured form for subsequent advanced processing and dedicated maintenance service recommendation algorithms.

[0060] This information includes a detailed description of the fault, an analysis of potential impacts, and optimized recommended repair actions.

[0061] The specific process of data collection and preprocessing in the fault detection algorithm is as follows:

[0062] B1, firstly, adopt data integration technology to systematically collect and organize the historical data of rail transit vehicles;

[0063] These data include but are not limited to real-time monitoring data from vehicle sensors, detailed maintenance records, and operation logs, which together form the basis for algorithm training and optimization;

[0064] B2, then preprocess the data, and in the data preprocessing stage, use data cleaning technology to eliminate redundancy, errors and outliers in the data to ensure the accuracy and reliability of the data;

[0065] B3, then use the denoising algorithm to reduce the noise interference in the sensor data and improve the signal-to-noise ratio of the data;

[0066] B4. Finally, through data standardization, we ensure that data from different sources and scales can be compared and analyzed under the same framework, thereby ensuring the quality and consistency of the calculation data.

[0067] The method for constructing the knowledge graph is as follows:

[0068] C1, First, extensively collect relevant data on rail transit vehicles to ensure the comprehensiveness and diversity of the data. These data can come from real-time sensor data streams, organized maintenance record databases, and technical documents and parameter tables obtained directly from manufacturers;

[0069] The collected data includes but is not limited to detailed specifications of vehicle parts, historical failure mode records, maintenance logs, and technical documents and parameters provided by manufacturers. By integrating these data sources, it can provide solid and rich basic data support for the construction of the knowledge graph.

[0070] C2, after obtaining the original data, execute the data annotation process, accurately classify each data according to its characteristics and relevance through data annotation technology, and build the corresponding data model to provide a structured data foundation for the subsequent knowledge graph construction and ensure the logic and consistency of the data;

[0071] The data annotation content includes but is not limited to the overall hierarchical structure of the vehicle, the detailed attributes of each component, specific failure modes and their characteristic descriptions. These structured data will provide strong support for the construction of the knowledge graph and promote the precise application of the knowledge graph in the field of fault diagnosis.

[0072] C3, finally, the knowledge graph architecture is designed through the constructed data model, and the hierarchical structure of rail transit vehicles and the correlation between internal components are fully considered during the design to ensure that the knowledge graph can fully and accurately reflect the complexity and dynamics of the vehicle system;

[0073] C4, finally, accurately define the entities (such as vehicle parts, fault types, maintenance event records), relationships (such as the association between parts and faults, the mapping relationship between faults and maintenance event records), and attributes (such as the material properties of parts, the impact assessment of faults) in the graph to complete the graph model design.

[0074] By constructing a comprehensive and detailed rail transit vehicle knowledge graph that covers key data such as detailed information on vehicle components, failure mode classification, and maintenance history records, a structured knowledge system has been formed, providing a solid data foundation to support more accurate and efficient diagnostic analysis of rail transit vehicle failures.

[0075] Among them, the real-time update method of the knowledge graph is:

[0076] D1, first build a refined data interface and automatic crawling mechanism to collect real-time data efficiently and automatically;

[0077] Specifically, it involves real-time synchronization of vehicle sensor data, periodic retrieval of maintenance record systems, and instant capture of the latest update information released by manufacturers. This comprehensive data capture strategy ensures the timeliness and accuracy of information, providing a solid foundation for subsequent data analysis and decision support;

[0078] D2, the captured real-time data is cleaned (to eliminate noise and outliers), deduplicated (to ensure data uniqueness), and missing information is supplemented (through interpolation, prediction or external data sources), so that these data can be efficiently integrated with the existing data in the knowledge graph;

[0079] These operations ensure the data integrity, consistency, and accuracy of the knowledge graph, providing reliable data support for subsequent data analysis and fault diagnosis;

[0080] D3, through database storage technology, accurately associates the processed and updated data with the knowledge graph, thereby achieving real-time updating of the knowledge graph.

[0081] This process uses an efficient data synchronization mechanism to ensure that the knowledge graph can accurately reflect the actual status of rail transit vehicles, including their latest component information, fault records, and maintenance dynamics. Through this real-time update method, it provides strong data support for fault diagnosis, preventive maintenance and vehicle management.

[0082] Among them, in the construction and update module of the knowledge graph, when constructing the knowledge graph, it is necessary to associate the complex hierarchical structure of the rail transit vehicle, including the hierarchical relationship from the vehicle as a whole to each component, and by defining the interaction and dependency relationship between them, the knowledge graph is made more instructive and practical in the actual application of rail transit vehicles;

[0083] In the construction and update module of the knowledge graph, an efficient update mechanism is used to ensure that the knowledge graph can quickly capture and integrate changes in new data during updating, thereby improving its dynamic response capability.

[0084] This design enables the knowledge graph to maintain a high degree of real-time performance, thereby more accurately supporting fault diagnosis of rail transit vehicles.

[0085] The specific process of the maintenance service recommendation algorithm is as follows:

[0086] E1, when the vehicle system detects an abnormal condition, the system fault detection mechanism will be triggered immediately, thus starting a detailed fault detection process;

[0087] During this process, the system will accurately collect comprehensive information related to the fault, including but not limited to key data such as the specific type of fault, the exact location of the component where the fault occurred, and the severity of the fault. The collection of this data provides a solid data foundation for subsequent fault diagnosis and maintenance service recommendations;

[0088] E2, accurately maps the detected fault information into the constructed knowledge graph structure, and then deeply analyzes the potential fault mode and possible impact range;

[0089] During this process, the algorithm will use the correlation relationships in the graph to extract key information such as component information closely related to the fault, historical failure mode data, and maintenance records, thereby providing a comprehensive and solid foundation for subsequent maintenance service recommendations. This knowledge graph-based analysis method not only improves the accuracy and efficiency of fault analysis, but also ensures the pertinence and effectiveness of maintenance service recommendations.

[0090] E3, at this time, the maintenance service recommendation algorithm directly accesses the vehicle's maintenance record database through an efficient data interface to obtain detailed historical maintenance records of the vehicle, and accurately identifies the processing mode and experience of similar faults in the past, thereby providing a more accurate and effective recommendation plan for the current fault repair;

[0091] These historical maintenance records cover key information such as the type of vehicle failure, maintenance plan, maintenance operation duration, and the qualifications and experience of professionals involved in the maintenance. This data-driven approach based on historical maintenance records significantly improves the accuracy and reliability of maintenance recommendations and provides more scientific decision-making support for vehicle maintenance.

[0092] E4, at the same time, the algorithm will obtain comprehensive information of the maintenance team in real time for accurate capture and integration, so that it can provide more accurate and efficient personnel allocation suggestions for fault maintenance tasks based on the actual capabilities and current status of the maintenance personnel;

[0093] This comprehensive information includes the maintenance personnel's professional skill level, current work status (such as idle, busy, offline, etc.), and real-time geographic location. Such key data can be accurately captured and integrated through advanced personnel positioning systems (such as GPS positioning, RFID identification, etc.) and work management systems (such as ERP, CRM, etc.), thereby improving the overall efficiency of maintenance services;

[0094] E5, the final algorithm performs a series of evaluation operations to accurately determine the feasibility of different maintenance solutions by comprehensively considering the fault details, the in-depth analysis results of the knowledge graph, the historical maintenance records, and the real-time maintenance team status;

[0095] This evaluation process involves consideration of multiple factors, including but not limited to the expected maintenance time, cost-benefit analysis, the matching degree between maintenance personnel skills and task requirements, and the availability of required replacement parts (including inventory status, supplier delivery cycle, etc.). Through this series of refined evaluations, the algorithm can ensure that the recommended maintenance plan is both efficient and economical, while meeting the dual optimization of skills and resources, providing optimal decision support for vehicle maintenance;

[0096] E6, at the same time, through advanced optimization algorithms such as Genetic Algorithm (GA) and Simulated Annealing (SA), potential maintenance solutions are arranged, combined and iteratively optimized to find the global optimal solution;

[0097] The core of this step is to seek the best balance between multiple conflicting factors, including but not limited to maintenance time, cost, personnel skills, parts availability, etc., in order to significantly improve maintenance efficiency while effectively reducing operating costs. Through the application of these optimization algorithms, it can ensure that the selected maintenance plan is not only technically feasible, but also optimal in terms of economy and efficiency, providing strong decision-making support for the organization;

[0098] E7,Finally, the algorithm will generate a detailed and optimized maintenance service recommendation plan, which includes a series of specific maintenance operation steps, estimated completion time, and detailed information of professional maintenance personnel assigned for each task.

[0099] This recommendation result will be presented accurately to relevant maintenance executors and decision-makers in a standardized report format through an efficient data transmission system to ensure that maintenance tasks can be efficiently executed according to the most optimized plan.

[0100] Among them, the service recommendation system collects detailed feedback information in real time through a data capture mechanism during the maintenance process. The collected high-quality data will be used to dynamically adjust the algorithm, thereby continuously optimizing the recommendation model and improving the system's adaptability and prediction accuracy.

[0101] These feedbacks cover the time consumed in actual maintenance operations, detailed maintenance cost breakdowns, and actual effect evaluations after fault resolution. Through this iterative process, the system can continuously learn and evolve to provide more accurate and efficient maintenance service recommendations.

[0102] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0103] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph, characterized by: The service recommendation system is composed of a rail transit vehicle fault detection module, a knowledge graph construction and update module, and a maintenance service recommendation module; The rail transit vehicle fault detection module is implemented based on a fault detection algorithm, the knowledge graph construction and update module includes the construction of a knowledge graph and the real-time update of a knowledge graph, and the maintenance service recommendation module is implemented based on a maintenance service recommendation algorithm.

2. The rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph according to claim 1 is characterized by: The fault detection algorithm includes data collection and preprocessing, feature extraction, deep learning model training, continuous learning and updating, real-time monitoring and fault identification and fault diagnosis result output, and the specific process of the fault detection algorithm is as follows: A1, first collect and organize the historical data of rail transit vehicles, and standardize the data; A2, through the efficient processing of huge sensor data sets through deep learning technology, and then accurately extracting significant features; A3, then, make full use of historical sensor data to train the deep learning model, and finally build an accurate vehicle status monitoring and potential fault prediction model through repeated iteration and optimization; A4, then introduces an adaptive continuous learning and updating mechanism. During the actual operation of the vehicle, the deep learning model is incrementally learned online using real-time streaming monitoring data to dynamically optimize the model parameters. A5, at this point the model has the ability to continuously and in real time monitor the vehicle sensor data stream, and based on the real-time input data, the built-in algorithm can efficiently identify abnormal patterns, thereby accurately indicating the potential source of the fault; A6, the algorithm will output the accurately identified failure mode and its related information in a structured form for subsequent advanced processing and dedicated maintenance service recommendation algorithms.

3. The rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph according to claim 2 is characterized by: The specific process of data collection and preprocessing in the fault detection algorithm is as follows: B1, firstly, adopt data integration technology to systematically collect and organize the historical data of rail transit vehicles; B2, then preprocess the data, and in the data preprocessing stage, use data cleaning technology to eliminate redundancy, errors and outliers in the data to ensure the accuracy and reliability of the data; B3, then use the denoising algorithm to reduce the noise interference in the sensor data and improve the signal-to-noise ratio of the data; B4. Finally, through data standardization, we ensure that data from different sources and scales can be compared and analyzed under the same framework, thereby ensuring the quality and consistency of the calculation data.

4. The rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph according to claim 1 is characterized by: The method for constructing the knowledge graph is as follows: C1, First, extensively collect relevant data on rail transit vehicles to ensure the comprehensiveness and diversity of the data. These data can come from real-time sensor data streams, organized maintenance record databases, and technical documents and parameter tables obtained directly from manufacturers; C2, after obtaining the original data, execute the data annotation process, accurately classify each data according to its characteristics and relevance through data annotation technology, and build the corresponding data model to provide a structured data foundation for the subsequent knowledge graph construction and ensure the logic and consistency of the data; C3, finally, the knowledge graph architecture is designed through the constructed data model, and the hierarchical structure of rail transit vehicles and the correlation between internal components are fully considered during the design to ensure that the knowledge graph can fully and accurately reflect the complexity and dynamics of the vehicle system; C4, finally, accurately define the entities, relationships, and attributes in the graph to complete the graph model design.

5. The rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph according to claim 1 is characterized by: The real-time updating method of the knowledge graph is: D1, first build a refined data interface and automatic crawling mechanism to collect real-time data efficiently and automatically; D2, cleans, removes duplications, and supplements missing information of the captured real-time data, so that the data can be efficiently integrated with the existing data in the knowledge graph; D3, through database storage technology, accurately associates the processed and updated data with the knowledge graph, thereby achieving real-time updating of the knowledge graph.

6. The rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph according to claim 1 is characterized by: In the construction and update module of the knowledge graph, when constructing the knowledge graph, it is necessary to associate the complex hierarchical structure of the rail transit vehicle, including the hierarchical relationship from the vehicle as a whole to each component, and by defining the interaction and dependency relationship between them, the knowledge graph is made more instructive and practical in the actual application of rail transit vehicles; In the construction and update module of the knowledge graph, an efficient update mechanism is used to ensure that the knowledge graph can quickly capture and integrate changes in new data during updating, thereby improving its dynamic response capability.

7. The rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph according to claim 1 is characterized by: The specific process of the maintenance service recommendation algorithm is as follows: E1, when the vehicle system detects an abnormal condition, the system fault detection mechanism will be triggered immediately, thus starting a detailed fault detection process; E2, accurately maps the detected fault information into the constructed knowledge graph structure, and then deeply analyzes the potential fault mode and possible impact range; E3, at this time, the maintenance service recommendation algorithm directly accesses the vehicle's maintenance record database through an efficient data interface to obtain detailed historical maintenance records of the vehicle, and accurately identifies the processing mode and experience of similar faults in the past, thereby providing a more accurate and effective recommendation plan for the current fault repair; E4, at the same time, the algorithm will obtain comprehensive information of the maintenance team in real time for accurate capture and integration, so that it can provide more accurate and efficient personnel allocation suggestions for fault maintenance tasks based on the actual capabilities and current status of the maintenance personnel; E5, the final algorithm performs a series of evaluation operations to accurately determine the feasibility of different maintenance solutions by comprehensively considering the fault details, the in-depth analysis results of the knowledge graph, the historical maintenance records, and the real-time maintenance team status; E6, at the same time, through advanced optimization algorithms, potential maintenance solutions are arranged, combined and iteratively optimized to find the global optimal solution; E7,Finally, the algorithm will generate a detailed and optimized maintenance service recommendation plan, which includes a series of specific maintenance operation steps, estimated completion time, and detailed information of professional maintenance personnel assigned for each task.

8. The rail transit vehicle fault intelligent diagnosis and maintenance service recommendation system based on knowledge graph according to claim 1 is characterized by: The service recommendation system collects detailed feedback information in real time through a data capture mechanism during the maintenance process. The collected high-quality data will be used to dynamically adjust the algorithm, thereby continuously optimizing the recommendation model and improving the system's adaptability and prediction accuracy.

Images