Train control equipment operation and maintenance method and train control equipment operation and maintenance device

Abstract

The invention provides a train control equipment operation and maintenance method and a train control equipment operation and maintenance device, and relates to the technical field of computers. The method comprises the following steps: initializing a digital twinborn model of the train control equipment based on equipment information of the train control equipment; the method comprises the following steps: collecting operation and maintenance data of train control equipment in real time, performing entity recognition processing on the operation and maintenance data, extracting entity knowledge information, performing fusion processing on the entity knowledge information, and constructing an operation and maintenance knowledge graph of the train control equipment; and importing the operation and maintenance knowledge graph into a database of a digital twin model, simulating an actual operation state of the train control equipment through the digital twin model, predicting a future operation state of the train control equipment, and determining an operation and maintenance strategy of the train control equipment based on the future operation state. Real-time monitoring and remote diagnosis of the equipment are achieved through the digital twinborn technology, operation and maintenance data are managed and analyzed through the knowledge graph technology, intelligent operation and maintenance decision support is provided, and then the reliability and safety of the train control equipment in the complex environment are improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a train control equipment operation and maintenance method and a train control equipment operation and maintenance device. Background Art

[0002] In high-altitude and cold areas, since railway train control equipment is exposed to extreme environments such as high altitude, severe cold, and large temperature difference between day and night for a long time, the concealment and suddenness of equipment failures increase. With the growing demand for operation and maintenance, it is very important to operate and maintain the train control equipment to ensure its reliability and safety.

[0003] Existing operation and maintenance methods include manual inspection and maintenance, static analysis relying on offline data, and using accumulated historical operation data to build a fault classification model to judge the equipment status. Since the above methods are still at the static analysis stage of offline data, there is a gap in the multi-level and in-depth operation status monitoring of train control equipment, and an effective connection between equipment status and operation and maintenance decisions has not yet been established. As a result, it is difficult to ensure the reliability and safety of train control equipment in complex environments, resulting in low operation reliability and maintenance efficiency of train control equipment in extreme environments. Summary of the invention

[0004] The purpose of the present application is to provide a train control equipment operation and maintenance method and a train control equipment operation and maintenance device to address the deficiencies in the above-mentioned prior art, so as to solve the problem that the existing operation and maintenance methods are difficult to ensure the reliability and safety of the train control equipment when facing complex environments.

[0005] To achieve the above purpose, the technical solution adopted in the embodiment of the present application is as follows:

[0006] In a first aspect, an embodiment of the present application provides a train control equipment operation and maintenance method, the method comprising:

[0007] Initialize the digital twin model of the train control device based on the device information of the train control device;

[0008] Collecting operation and maintenance data of the train control equipment in real time, performing entity recognition processing on the operation and maintenance data, extracting entity knowledge information from the operation and maintenance data, and fusing the entity knowledge information to construct an operation and maintenance knowledge graph of the train control equipment, wherein the entity knowledge information includes: equipment name, fault type, and maintenance plan;

[0009] The operation and maintenance knowledge graph is imported into the database of the digital twin model, the actual operation status of the train control device is simulated through the digital twin model, the future operation status of the train control device is predicted, and the operation and maintenance strategy of the train control device is determined based on the future operation status of the virtual model of the train control device.

[0010] As a possible implementation manner, simulating the actual operating state of the train control device by using the digital twin model and predicting the future operating state of the train control device includes:

[0011] Establishing an association relationship between the operation and maintenance knowledge graph and the real-time status data of the train control device through the digital twin model, simulating the operating status of the train control device based on the association relationship, and obtaining real-time operating data;

[0012] Based on the real-time operation data and the historical operation and maintenance data of the train control device, the future operation status of the train control device is predicted.

[0013] As a possible implementation manner, determining the operation and maintenance strategy of the train control device based on the future operation state of the virtual model of the train control device includes:

[0014] Based on the future operating state of the virtual model of the train control device, calling the fault diagnosis information in the operation and maintenance knowledge graph;

[0015] Determining the fault type according to the fault diagnosis information;

[0016] Based on the fault type, an operation and maintenance strategy of the train control device is generated.

[0017] As a possible implementation method, the operation and maintenance data includes: operation log, operation abnormality analysis report, vehicle equipment failure information, and maintenance disposal plan;

[0018] After collecting the operation and maintenance data of the train control device, the method further includes:

[0019] Performing data cleaning on the operation and maintenance data to obtain processed operation and maintenance data;

[0020] The performing entity recognition processing on the operation and maintenance data to extract entity knowledge information in the operation and maintenance data includes:

[0021] Perform entity recognition processing on the processed operation and maintenance data to extract the entity knowledge information.

[0022] As a possible implementation manner, performing entity recognition processing on the processed operation and maintenance data to extract the entity knowledge information includes:

[0023] Performing labeling processing on the processed operation and maintenance data to obtain labeled operation and maintenance data;

[0024] Inputting the annotated operation and maintenance data into a pre-trained deep learning model, and converting the annotated operation and maintenance data into an operation and maintenance vector based on the deep learning model;

[0025] Extracting context feature information from the operation and maintenance vector;

[0026] A tag sequence corresponding to the context feature information is determined, and based on the tag sequence, the entity knowledge information is extracted.

[0027] As a possible implementation manner, the fusing of the entity knowledge information to construct an operation and maintenance knowledge graph of the train control device includes:

[0028] Building an entity dictionary based on the entity knowledge information;

[0029] For each entity in the entity dictionary, determining a similarity between the entity and any remaining entity in the entity dictionary except the entity;

[0030] Based on the similarity, determining whether to perform fusion processing on the entity and the remaining entities;

[0031] Based on the fused entities, the operation and maintenance knowledge graph is constructed.

[0032] As a possible implementation manner, determining whether to merge the entity and the remaining entities based on the similarity includes:

[0033] The similarity is compared with a preset threshold. If the similarity is greater than the preset threshold, it is determined to fuse the entity and the remaining entities into one entity, and to delete duplicate entities in the entity dictionary. The duplicate entity is any entity between the entity and the remaining entities.

[0034] As a possible implementation manner, before fusing the entity knowledge information and constructing the operation and maintenance knowledge graph of the train control device, the method further includes:

[0035] The entity knowledge information is input into a pre-trained hybrid network model, and relationship extraction is performed on the entity knowledge information based on the hybrid network model to determine entity association relationships, wherein the entity association relationships include: an association relationship between a device name and a fault type, and an association relationship between a fault type and a maintenance plan.

[0036] As a possible implementation method, the operation and maintenance knowledge graph is constructed based on the fused entity, including:

[0037] Based on the fused entities and the entity association relationships, the operation and maintenance knowledge graph is constructed.

[0038] In a second aspect, an embodiment of the present application provides a train control equipment operation and maintenance device, the device comprising:

[0039] An initialization module is used to initialize the digital twin model of the train control device based on the device information of the train control device;

[0040] An entity processing module is used to collect the operation and maintenance data of the train control equipment in real time, perform entity recognition processing on the operation and maintenance data, extract entity knowledge information from the operation and maintenance data, and perform fusion processing on the entity knowledge information to construct an operation and maintenance knowledge graph of the train control equipment, wherein the entity knowledge information includes: equipment name, fault type, and maintenance plan;

[0041] A simulation prediction module is used to import the operation and maintenance knowledge graph into the database of the digital twin model, simulate the actual operation status of the train control equipment through the digital twin model, predict the future operation status of the train control equipment, and determine the operation and maintenance strategy of the train control equipment based on the future operation status of the virtual model of the train control equipment.

[0042] As a possible implementation, the simulation prediction module is specifically used for:

[0043] Establishing an association relationship between the operation and maintenance knowledge graph and the real-time status data of the train control device through the digital twin model, simulating the operating status of the train control device based on the association relationship, and obtaining real-time operating data;

[0044] Based on the real-time operation data and the historical operation and maintenance data of the train control device, the future operation status of the train control device is predicted.

[0045] As a possible implementation, the simulation prediction module is specifically used for:

[0046] Based on the future operating state of the virtual model of the train control device, calling the fault diagnosis information in the operation and maintenance knowledge graph;

[0047] Determining the fault type according to the fault diagnosis information;

[0048] Based on the fault type, an operation and maintenance strategy of the train control device is generated.

[0049] As a possible implementation, the operation and maintenance data includes: operation log, operation abnormality analysis report, vehicle equipment failure information, and maintenance disposal plan; the entity processing module is specifically used to:

[0050] Performing data cleaning on the operation and maintenance data to obtain processed operation and maintenance data;

[0051] Perform entity recognition processing on the processed operation and maintenance data to extract the entity knowledge information.

[0052] As a possible implementation, the entity processing module is specifically used to:

[0053] Performing labeling processing on the processed operation and maintenance data to obtain labeled operation and maintenance data;

[0054] Inputting the annotated operation and maintenance data into a pre-trained deep learning model, and converting the annotated operation and maintenance data into an operation and maintenance vector based on the deep learning model;

[0055] Extracting context feature information from the operation and maintenance vector;

[0056] A tag sequence corresponding to the context feature information is determined, and based on the tag sequence, the entity knowledge information is extracted.

[0057] As a possible implementation, the entity processing module is specifically used to:

[0058] Building an entity dictionary based on the entity knowledge information;

[0059] For each entity in the entity dictionary, determining a similarity between the entity and any remaining entity in the entity dictionary except the entity;

[0060] Based on the similarity, determining whether to perform fusion processing on the entity and the remaining entities;

[0061] Based on the fused entities, the operation and maintenance knowledge graph is constructed.

[0062] As a possible implementation, the entity processing module is specifically used to:

[0063] The similarity is compared with a preset threshold. If the similarity is greater than the preset threshold, it is determined to fuse the entity and the remaining entities into one entity, and to delete duplicate entities in the entity dictionary. The duplicate entity is any entity between the entity and the remaining entities.

[0064] As a possible implementation manner, the entity processing module is further used to:

[0065] The entity knowledge information is input into a pre-trained hybrid network model, and relationship extraction is performed on the entity knowledge information based on the hybrid network model to determine entity association relationships, wherein the entity association relationships include: an association relationship between a device name and a fault type, and an association relationship between a fault type and a maintenance plan.

[0066] As a possible implementation, the entity processing module is specifically used to:

[0067] Based on the fused entities and the entity association relationships, the operation and maintenance knowledge graph is constructed.

[0068] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium through the bus, and the processor executes the machine-readable instructions to perform the steps of the train control equipment operation and maintenance method as described in any one of the first aspects above.

[0069] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the train control equipment operation and maintenance method as described in any one of the first aspects above are executed.

[0070] According to the train control equipment operation and maintenance method and train control equipment operation and maintenance device of the embodiment of the present application, based on the equipment information of the train control equipment, the digital twin model of the train control equipment is initialized; the operation and maintenance data of the train control equipment is collected in real time, entity recognition processing is performed on the operation and maintenance data, entity knowledge information therein is extracted, and the entity knowledge information is fused and processed to construct the operation and maintenance knowledge graph of the train control equipment; the operation and maintenance knowledge graph is imported into the database of the digital twin model, the actual operation state of the train control equipment is simulated by the digital twin model, its future operation state is predicted, and the operation and maintenance strategy of the train control equipment is determined based on the future operation state of the virtual model. According to the embodiment of the present application, by collecting the operation and maintenance data of the train control equipment in real time, performing entity recognition processing on it, and extracting entity knowledge information, the transition from static analysis to dynamic monitoring is realized, and through real-time data collection, the abnormal situation of the equipment can be discovered in time and responded quickly to avoid the accumulation of problems caused by data lag. By fusion processing of the entity knowledge information in the operation and maintenance data, the operation and maintenance knowledge graph of the train control equipment is constructed, and the knowledge graph technology is used to deeply analyze the operation state of the equipment, and realize multi-level and multi-dimensional monitoring of the operation state of the equipment. The operation and maintenance knowledge graph is then imported into the database of the digital twin model. By simulating the operating status of the equipment under different environmental conditions, the performance of the train control equipment in extreme environments can be predicted and evaluated. Based on these prediction results, a reasonable operation and maintenance strategy can be formulated, thereby improving the reliability and safety of the train control equipment in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0072] Figure 1 A flow chart of a train control equipment operation and maintenance method provided in an embodiment of the present application is shown;

[0073] Figure 2 A schematic diagram of a process for extracting entity knowledge information provided by an embodiment of the present application is shown;

[0074] Figure 3 A schematic diagram of a process for constructing an operation and maintenance knowledge graph provided in an embodiment of the present application is shown;

[0075] Figure 4 A schematic diagram of an operation and maintenance process provided by an embodiment of the present application is shown;

[0076] Figure 5 A schematic diagram of the structure of a train control equipment operation and maintenance device provided in an embodiment of the present application is shown;

[0077] Figure 6 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0078] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart under the guidance of the content of the present application, or remove one or more operations from the flowchart.

[0079] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0080] In order to enable those skilled in the art to use the contents of this application, the following implementation method is given in combination with the specific application scenario "intelligent operation and maintenance of train control equipment based on digital twins and knowledge graphs". For those skilled in the art, the general principles defined here can be applied to other embodiments and application scenarios without departing from the spirit and scope of this application. Although this application is mainly described around the operation and maintenance method of train control equipment, it should be understood that this is only an exemplary embodiment.

[0081] It should be noted that the term "comprising" will be used in the embodiments of the present application to indicate the existence of the features declared thereafter, but does not exclude the addition of other features.

[0082] The harsh environment of the western cold region, such as high altitude, severe cold and large temperature difference between day and night, seriously affects the normal operation and maintenance of railway train control equipment. The traditional operation and maintenance management method is inefficient under complex environmental conditions, and it is difficult to ensure the reliability and safety of the equipment. To solve this problem, this application is based on digital twin technology and knowledge graph technology. By building a virtual model of the train control equipment, it can realize real-time monitoring and data collection of the equipment operation status, and timely discover and warn potential faults. At the same time, build an equipment operation and maintenance knowledge base, systematically manage equipment failure modes, maintenance records and operation and maintenance specifications, provide intelligent diagnosis and maintenance suggestions, and improve operation and maintenance efficiency and accuracy. By simulating the operating status of equipment under different environmental conditions, optimize equipment design and maintenance strategies, and improve the environmental adaptability and stability of equipment. In addition, build a unified data integration platform to centrally manage and analyze equipment operation data, environmental data and operation and maintenance data, use big data technology for fault prediction and trend analysis, provide scientific decision support, ensure high-speed railway driving safety, and improve the reliability and efficiency of railway transportation.

[0083] Figure 1 The following is a flow chart of a train control equipment operation and maintenance method provided by an embodiment of the present application. Figure 1 As shown, the method specifically comprises the following steps:

[0084] S101. Initialize a digital twin model of the train control device based on the device information of the train control device.

[0085] Optionally, the equipment information includes equipment type parameters such as model, specification, and operation status of the train control equipment. By establishing a virtual equipment model (Building Information Modeling, BIM) of the train control equipment in the building information modeling tool Revit, and adding the above equipment information in the virtual equipment model BIM, it is ensured that the virtual equipment model BIM contains not only the geometric shape of the train control equipment, but also various attribute data of the train control equipment.

[0086] Furthermore, after the virtual equipment model BIM is created, the virtual equipment model BIM is imported into a 3D modeling tool (3D Studio Max, 3ds Max) for high-precision rendering to obtain a more realistic visual effect and improve the realism of the virtual equipment model BIM. After the rendering is completed, the virtual equipment model BIM is converted into an FBX format model file supported by Unity3D. The FBX format is a universal 3D file format that can easily transfer 3D models between different software platforms. On this basis, a virtual scene is built in Unity3D, and the converted FBX format model file is imported to program the train control system.

[0087] In the actual physical system space, various sensors are deployed on the train control equipment, such as temperature sensors, pressure sensors, vibration sensors, etc., which can be used to detect various status parameters of the train control equipment. In addition, the building information modeling tool Revit, 3D modeling tool 3ds Max and Unity3D are integrated to create a complete train control equipment information database, which not only contains the detailed parameters of the train control equipment (such as model, specification, operating status, etc.), but also includes the virtual model and operating status information of the train control equipment, which can be used to monitor the operating status of the equipment in real time, perform fault diagnosis, and assist decision-making and support maintenance activities.

[0088] S102. Collect the operation and maintenance data of the train control equipment in real time, perform entity recognition processing on the operation and maintenance data, extract entity knowledge information from the operation and maintenance data, and fuse the entity knowledge information to construct an operation and maintenance knowledge graph of the train control equipment.

[0089] Optionally, entity knowledge information includes but is not limited to: equipment name, fault type, and maintenance plan. Operation and maintenance data includes but is not limited to: operation log, operation abnormality analysis report, vehicle-mounted equipment fault information, and maintenance disposal plan. Among them, the equipment name includes identification information such as the model and serial number of the equipment, the fault type refers to the specific fault type that may occur in the train control equipment, and the maintenance plan refers to the maintenance method and steps for a specific fault type. The operation log refers to information recording the operating status of the train control equipment, such as the working parameters and operation records of the equipment, etc. The operation abnormality analysis report refers to an analysis report generated when an abnormality occurs in the train control equipment, including the cause of the abnormality, the scope of impact, etc. The vehicle-mounted equipment fault information refers to recording the fault information of the vehicle-mounted equipment, such as fault code, fault description, etc. The maintenance disposal plan refers to recording the maintenance plan, maintenance steps, required materials, etc. used in the equipment maintenance process.

[0090] Optionally, by performing natural language processing on the operation and maintenance data collected in real time, key objects in the operation and maintenance data are identified, including but not limited to entity knowledge information such as equipment name, fault type and maintenance plan. The entity knowledge information extracted from the operation and maintenance data is fused and processed to establish associations between entity knowledge, such as the association between equipment name and fault type, the association between fault type and maintenance plan, and the association between maintenance plan and other relevant factors (such as time, location, operator, etc.), to form an operation and maintenance knowledge graph. The nodes in the operation and maintenance knowledge graph represent entities (such as equipment name, fault type, maintenance plan), and the edges represent the relationships between entities.

[0091] S103. Import the operation and maintenance knowledge graph into the database of the digital twin model, simulate the actual operation status of the train control equipment through the digital twin model, predict the future operation status of the train control equipment, and determine the operation and maintenance strategy of the train control equipment based on the future operation status of the virtual model of the train control equipment.

[0092] Optionally, the data in the operation and maintenance knowledge graph is converted into a format suitable for the database of the digital twin model and imported into the database of the digital twin model. The status data of the train control equipment is collected in real time using monitoring equipment such as sensors, and an association relationship between the operation and maintenance knowledge graph and the status data is established. The status of the virtual equipment model BIM is updated based on the real-time collected status data to reflect the current operation status of the train control equipment.

[0093] Optionally, the operating status of the train control equipment is simulated in real time in the digital twin environment, trend analysis is performed based on real-time operating data and historical operation and maintenance data, the future operating status of the train control equipment is predicted, potential failure modes of the train control equipment are identified, possible problems are warned in advance, and the remaining life of the equipment is predicted based on the frequency of use and historical maintenance records of the equipment. Furthermore, based on the prediction results, intelligent diagnosis is performed using the operation and maintenance knowledge graph to identify the root causes that may lead to failures, and targeted maintenance suggestions are provided based on the diagnosis results and historical maintenance records to optimize the design and maintenance strategies of the train control equipment and ensure the efficient operation of the train control equipment.

[0094] Based on this, according to the train control equipment operation and maintenance method provided in the embodiment of the present application, by real-time collection of the operation and maintenance data of the train control equipment, and entity recognition processing, and extracting entity knowledge information, the transformation from static analysis to dynamic monitoring is realized, and through real-time data collection, abnormal conditions of the equipment can be discovered in time, and rapid responses can be made to avoid the accumulation of problems caused by data lags. By fusing the entity knowledge information in the operation and maintenance data, constructing the operation and maintenance knowledge graph of the train control equipment, and using the knowledge graph technology, the operating status of the equipment can be deeply analyzed, and multi-level and multi-dimensional monitoring of the operating status of the equipment can be realized. Then, the operation and maintenance knowledge graph is imported into the database of the digital twin model, and by simulating the equipment operating status under different environmental conditions, the performance of the train control equipment in extreme environments is predicted and evaluated. Based on these prediction results, reasonable operation and maintenance strategies can be formulated, thereby improving the reliability and safety of the train control equipment in complex environments.

[0095] As a possible implementation manner, after collecting the operation and maintenance data of the train control equipment, the method further includes: performing data cleaning on the operation and maintenance data to obtain processed operation and maintenance data.

[0096] Optionally, data cleaning is to process the collected operation and maintenance data to remove noise in the operation and maintenance data, such as duplicate records and format errors, to ensure the integrity and accuracy of the operation and maintenance data. Exemplarily, duplicate records are removed by checking whether there are duplicate records in the operation and maintenance data, and deleting or merging duplicate items. For example, multiple records of the same train control device at the same time can be merged into one record to ensure that data from different sources are consistent when describing the same event.

[0097] Exemplarily, by standardizing the formats of data fields in the operation and maintenance data, such as unifying the date and time formats, and correcting spelling errors, numerical errors, etc. in the operation and maintenance data, the consistency and readability of the data can be ensured. In addition, for missing data items in the operation and maintenance data, they can be filled in according to the context or statistical methods to ensure that the filled values ​​are reasonable and consistent with the actual situation. For outliers in the operation and maintenance data, statistical methods can be used to identify them, and determine whether to retain the outliers or replace them with reasonable values ​​according to the actual situation. Furthermore, after the data cleaning is completed, a train control equipment operation and maintenance corpus can be further constructed to cover the key knowledge points in the field, and the data in the corpus can be marked, such as marking entity information such as equipment name, fault type, and maintenance plan.

[0098] Based on this, the data quality of operation and maintenance data is improved by cleaning the data, so as to more accurately simulate and predict the operating status of equipment and formulate effective operation and maintenance strategies.

[0099] Furthermore, the above step S102 performs entity recognition processing on the operation and maintenance data to extract entity knowledge information in the operation and maintenance data, including: performing entity recognition processing on the processed operation and maintenance data to extract entity knowledge information.

[0100] Optionally, the processed operation and maintenance data is cleaned and preprocessed to remove noise such as duplicate records and format errors in the operation and maintenance data, thereby improving the integrity and consistency of the operation and maintenance data. On this basis, entity recognition processing is performed based on the processed operation and maintenance data, which significantly improves the accuracy and feasibility of data analysis.

[0101] Figure 2 The flowchart of a method for extracting entity knowledge information provided by an embodiment of the present application is shown. Figure 2 As shown, the above steps perform entity recognition processing on the processed operation and maintenance data to extract entity knowledge information, which specifically includes the following steps:

[0102] S201: labeling the processed operation and maintenance data to obtain labeled operation and maintenance data.

[0103] Exemplarily, part of the processed operation and maintenance data can be labeled, for example, entity knowledge information such as equipment name, fault type, and maintenance plan can be marked to form labeled operation and maintenance data with labeled tags. Specifically, first clarify the entity categories that need to be labeled, such as equipment name, fault type, maintenance plan, etc., and define the labeling rules corresponding to each category. For example, the labeling rule corresponding to the equipment name is the model and number of the equipment, the labeling rule corresponding to the fault type is to describe the specific problems or symptoms of the train control equipment, and the labeling rule corresponding to the maintenance plan is to describe the specific steps or measures to solve the problem. On this basis, the entities of each category are labeled based on the labeling rules corresponding to each category to obtain the labeled operation and maintenance data.

[0104] S202: Input the annotated operation and maintenance data into a pre-trained deep learning model, and convert the annotated operation and maintenance data into an operation and maintenance vector based on the deep learning model.

[0105] Exemplarily, the deep learning model can be a pre-trained model (Bidirectional Encoder Representations from Transformers, BERT), the annotated operation data is input into the pre-trained BERT model, the annotated operation data is encoded using the pre-trained BERT model, and the embedding vector of each word is generated, so that the text data corresponding to the annotated operation data is converted into a character-level embedding vector T, and the character-level embedding vector T is used as the operation vector. Among them, each element of the character-level embedding vector T is a vector, which represents the embedding representation of the corresponding word in the text.

[0106] S203: Extract context feature information from the operation and maintenance vector.

[0107] Exemplarily, the character-level embedding vector T is used as the output sequence of the BERT model, and the character-level embedding vector T is input into the bidirectional long short-term memory network (Bidirectional Long Short-Term Memory, BiLSTM). At this time, the character-level embedding vector T generated by the BERT model is used as the input sequence of the BiLSTM. The BiLSTM processes the input sequence through the LSTM layer in two directions (e.g., forward and reverse), thereby capturing the contextual information of the vocabulary in the text to achieve the extraction of contextual feature information, and obtaining the intermediate vector O, and the intermediate vector O is used as the label sequence output by the BiLSTM layer. Among them, the intermediate vector O contains the contextual feature information of each vocabulary, and the intermediate vector O can be used as the label sequence output by the BiLSTM layer for subsequent label prediction.

[0108] S204: Determine a tag sequence corresponding to the context feature information, and extract entity knowledge information based on the tag sequence.

[0109] Exemplarily, the label sequence output by the BiLSTM layer, that is, the above-mentioned intermediate vector O, is used as the input of the conditional random field (CRF), the scores of all possible label sequences are calculated, and the label sequence with the highest score is selected to extract entity knowledge information, such as equipment name, fault type, maintenance plan and other entity knowledge information.

[0110] Based on this, entity knowledge information is effectively extracted from unstructured operation and maintenance text data, and the relationship between entities is identified. This not only improves the quality of the data, enhances the semantic understanding ability of the model and improves the accuracy of entity recognition, but also forms a closed-loop optimization mechanism, which is convenient for improving the operation and maintenance efficiency and reliability of train control equipment.

[0111] Figure 3 A schematic diagram of a method for constructing an operation and maintenance knowledge graph provided in an embodiment of the present application is shown. As a possible implementation method, refer to Figure 3 As shown, the above step S102 performs fusion processing on entity knowledge information to construct an operation and maintenance knowledge graph of train control equipment, including:

[0112] S301. Construct an entity dictionary based on entity knowledge information.

[0113] Optionally, since operation and maintenance data is text data, there are differences in the way text data is recorded, and the descriptions of the same equipment, fault phenomenon, treatment measures, etc. are different, resulting in duplication of entity knowledge information obtained after entity extraction of operation and maintenance data. In order to avoid the construction of duplicate nodes in the operation and maintenance knowledge graph, an entity dictionary is constructed based on the entity knowledge information extracted from the operation and maintenance data of the train control equipment to perform entity matching in the entity dictionary, and the fused entities are updated to the operation and maintenance knowledge graph, and duplicate or similar entities are deleted, so as to effectively improve the quality of the operation and maintenance knowledge graph and ensure the accuracy and consistency of the content in the operation and maintenance knowledge graph.

[0114] Exemplarily, the entity dictionary usually contains known entity names, such as equipment name, fault type, maintenance plan, etc. Building an entity dictionary helps improve the accuracy and efficiency of entity recognition. By collecting entity knowledge information, forming a preliminary entity list, normalizing the entity list to ensure the consistency of entity names, and constructing the normalized entity list into a corpus, an entity dictionary is constructed.

[0115] S302: For each entity in the entity dictionary, determine the similarity between the entity and any remaining entity in the entity dictionary except the entity.

[0116] Exemplarily, the cosine similarity algorithm can be used to determine the similarity between any two entities in the entity dictionary. In order to calculate the similarity, the textual entity names need to be converted into numerical form, such as the vector representation of each entity. Cosine similarity measures the similarity by measuring the cosine value of the angle between two vectors. On this basis, once the vector representation of the entity is obtained, the cosine similarity algorithm can be used to calculate the similarity between the entities.

[0117] S303: Based on the similarity, determine whether to merge the entity with the remaining entities.

[0118] Optionally, the similarity is compared with a preset threshold. If the similarity is greater than the preset threshold, it is determined to fuse the entity and the remaining entities, merge the entity and the remaining entities into one entity, and delete the duplicate entity in the entity dictionary. The duplicate entity is any entity in the entity and the remaining entities. Exemplarily, the setting of the preset threshold can be determined according to the specific application scenario and requirements, and the value range is usually between 0 and 1, for example, it can be set to 0.8, which is used to determine whether two entities are similar enough to be regarded as the same entity. The similarity between the entity and any remaining entity is calculated based on the above cosine similarity algorithm, and a similarity matrix between all entities is constructed based on the calculated similarity, so as to use the similarity matrix for fusion processing. Specifically, if the similarity between two entities is higher than the preset threshold, they are merged into one entity, and the duplicate entity is deleted from the dictionary, and then the entity dictionary is updated according to the fusion result and the entity to be deleted. Specifically, the entity to be deleted is removed from the dictionary, and the merged entity is used as a new entity.

[0119] S304. Construct an operation and maintenance knowledge graph based on the fused entities.

[0120] Optionally, before the above-mentioned step S102 performs fusion processing on the entity knowledge information to construct the operation and maintenance knowledge graph of the train control equipment, the method also includes: inputting the entity knowledge information into a pre-trained hybrid network model, extracting relationships on the entity knowledge information based on the hybrid network model, and determining entity association relationships.

[0121] Optionally, the entity association relationship includes, but is not limited to: the association relationship between the equipment name and the fault type, and the association relationship between the fault type and the maintenance plan. Exemplarily, the hybrid network model can be a BERT-CNN model, and the extracted entity knowledge information such as the equipment name, fault type, and maintenance plan is input into the BERT-CNN model for relationship extraction to identify the association relationship between the equipment name and the fault type, the fault type and the maintenance plan, etc. In this process, the entity knowledge information such as the equipment name, fault type, and maintenance plan is first input into BERT for encoding to generate a word vector, and then the generated word vector is input into CNN for training, and a fully connected layer and a soft max layer are added to extract the relationship between the target entities.

[0122] Furthermore, based on the fused entities, an operation and maintenance knowledge graph is constructed, including: based on the fused entities and entity associations, an operation and maintenance knowledge graph is constructed. Exemplarily, after the extraction and knowledge fusion processing of the operation and maintenance data of the train control equipment, key data such as the train control equipment fault information record and the vehicle workshop equipment abnormal information statistical summary table are integrated into a structured and standardized operation and maintenance knowledge form, and the operation and maintenance knowledge graph of the train control equipment is constructed based on a two-way collaborative method. Specifically, first, a model layer is constructed in a top-down manner to complete the abstract expression of the element concept, and then a data layer is constructed in a bottom-up manner to obtain and fill in the standardized knowledge units. Finally, the operation and maintenance knowledge triples are stored in the Neo4j database, covering the module-level faults, component-level faults, cause analysis, equipment phenomena, treatment measures, maintenance measures and other element information of the train control equipment and the associations between them.

[0123] Based on this, the entities are fused according to similarity calculation and preset thresholds, and the entity dictionary is updated to reduce redundancy in the entity dictionary. In addition, the entities are fused to improve the accuracy of the entities, and then the fused entities are used to construct the operation and maintenance knowledge graph to ensure that the obtained operation and maintenance knowledge graph is more credible.

[0124] As a possible implementation method, the above-mentioned step S103 simulates the actual operating status of the train control equipment through a digital twin model, including: establishing an association relationship between the operation and maintenance knowledge graph and the real-time status data of the train control equipment through the digital twin model, simulating the operating status of the train control equipment based on the association relationship, and obtaining real-time operation data; based on the real-time operation data and the historical operation and maintenance data of the train control equipment, predicting the future operating status of the train control equipment.

[0125] Exemplarily, real-time status data is collected from the actual running train control equipment, including but not limited to the temperature, pressure, current, voltage, speed, etc. of the equipment, and the real-time status data is stored in the database of the digital twin model. In the digital twin model, the real-time status data is associated with the relevant information in the operation and maintenance knowledge graph through API or other interfaces. On this basis, the real-time status data is mapped to the virtual device in the digital twin model, and the status of the virtual device is updated in real time to keep the status of the virtual device consistent with the status of the actual device. The operation of the device is simulated in the digital twin environment to generate real-time operation data. Furthermore, the real-time operation data is fused with the historical operation and maintenance data to generate a comprehensive data set, and the comprehensive data set is trained using machine learning or deep learning models (such as LSTM, GRU and other models) to learn the trend of equipment status changes, and based on the trained model, the future operation status of the equipment is predicted, including possible failures, maintenance requirements, etc.

[0126] Based on this, the digital twin model is used to simulate the actual operating status of the train control equipment, and the future status of the equipment is predicted based on real-time operating data and historical operation and maintenance data. This can not only improve the efficiency and accuracy of equipment operation and maintenance, but also help detect potential faults in advance, thereby avoiding equipment downtime and accidents.

[0127] As a possible implementation method, the above-mentioned step S103 determines the operation and maintenance strategy of the train control equipment based on the future operating state of the virtual model of the train control equipment, including: calling the fault diagnosis information in the operation and maintenance knowledge graph based on the future operating state of the virtual model of the train control equipment; determining the fault type according to the fault diagnosis information; and generating the operation and maintenance strategy of the train control equipment based on the fault type.

[0128] Optionally, based on the integration of the operation and maintenance knowledge graph and the digital twin model, machine learning and artificial intelligence algorithms can be applied to formulate intelligent operation and maintenance strategies based on the operation and maintenance knowledge in the operation and maintenance knowledge graph and the real-time data of the digital twin model. Exemplarily, the modules prone to failure of the train control equipment are mainly concentrated in: on-board safety computer (Automatic Train Protection Control Unit, AT PCU), secure digital input and output unit (Vital Digital Input / Output, VDX), train interface unit (Train Interface Unit, TIU), wireless communication unit (Radio Transmission Module, RTM), balise transmission module (Balise Transmission Module, BTM), speed and distance unit (Speed ​​and Distance Unit, SDU) and driver machine interface (Driver Machine Interface, DMI), and the corresponding fault types are shown in Table 1 below:

[0129] Table 1 Fault types of train control equipment

[0130]

[0131] For example, when the sensor detects an abnormality in the train control equipment, the sensor data will be transmitted to the digital twin model in real time, triggering the anomaly detection algorithm. The digital twin model can call the fault diagnosis information in the knowledge graph in real time, match the corresponding fault type, and automatically generate a maintenance plan. It will also be fed back to the maintenance personnel of the physical system through the digital twin model to guide the maintenance personnel to perform fast and effective maintenance operations, thereby reducing the downtime of the train control equipment and improving the operating efficiency of the train control equipment.

[0132] For example, refer to Figure 4 As shown in the figure, the train control equipment operation and maintenance knowledge graph contains detailed operation and maintenance knowledge of the train control equipment, such as fault type, fault cause, maintenance plan and other information. The structured information in the knowledge graph is transmitted to the database of the digital twin model through the data pipeline, which not only enables the digital twin model to reflect the status of the physical system in real time, but also can self-optimize according to the latest operation and maintenance knowledge, thereby improving the accuracy of fault prediction and the effectiveness of maintenance plans, ensuring the reliability of equipment operation and the efficiency of maintenance.

[0133] Exemplarily, the operating status and fault information of the physical system of the train control equipment will be fed back to the operation and maintenance knowledge graph in real time to update the operation and maintenance knowledge base. For example, when a certain type of fault occurs frequently, the operation and maintenance knowledge graph will record the fault information and analyze its root cause and optimal treatment plan through machine learning algorithms. The updated operation and maintenance knowledge will also be transmitted to the digital twin model again to optimize its fault prediction and diagnosis capabilities. In this way, driven by real-time data, the virtual model can continuously update and optimize the operation and maintenance strategy, forming a closed-loop optimization mechanism to ensure the continuous improvement of the operation and maintenance strategy, thereby further improving the reliability and operation efficiency of the train control equipment.

[0134] Based on this, this application uses digital twin and knowledge graph technology to build a real-time, interactive, and collaborative intelligent operation and maintenance platform, which improves the operation and maintenance management level and intelligence of train control equipment. Specifically, digital twin technology is used to achieve real-time monitoring and remote diagnosis of equipment, and knowledge graph technology is used to manage and analyze operation and maintenance data, provide intelligent operation and maintenance decision support, and develop a visual monitoring and operation and maintenance management system that integrates monitoring management, asset management, and operation and maintenance management. Furthermore, through intelligent analysis and judgment, it is possible not only to timely discover and warn of potential faults, optimize equipment design and maintenance strategies, but also to improve the environmental adaptability and stability of equipment, reduce equipment maintenance costs, ensure the safety of high-speed railway driving, and thus improve the reliability and efficiency of railway transportation.

[0135] Based on the same inventive concept, a train control equipment operation and maintenance device corresponding to the train control equipment operation and maintenance method is also provided in the embodiment of the present application. Since the principle of solving the problem by the train control equipment operation and maintenance device in the embodiment of the present application is similar to the above-mentioned train control equipment operation and maintenance method in the embodiment of the present application, the implementation of the train control equipment operation and maintenance device can refer to the implementation of the train control equipment operation and maintenance method, and the repeated parts will not be repeated.

[0136] Reference Figure 5 As shown, it is a structural schematic diagram of a train control equipment operation and maintenance device provided in an embodiment of the present application. The train control equipment operation and maintenance device 500 includes: an initialization module 501, an entity processing module 502, and a simulation prediction module 503, wherein:

[0137] An initialization module 501 is used to initialize the digital twin model of the train control device based on the device information of the train control device;

[0138] The entity processing module 502 is used to collect the operation and maintenance data of the train control equipment in real time, perform entity recognition processing on the operation and maintenance data, extract entity knowledge information from the operation and maintenance data, and perform fusion processing on the entity knowledge information to construct the operation and maintenance knowledge graph of the train control equipment. The entity knowledge information includes: equipment name, fault type, and maintenance plan;

[0139] The simulation prediction module 503 is used to import the operation and maintenance knowledge graph into the database of the digital twin model, simulate the actual operation status of the train control equipment through the digital twin model, predict the future operation status of the train control equipment, and determine the operation and maintenance strategy of the train control equipment based on the future operation status of the virtual model of the train control equipment.

[0140] Based on this, according to the train control equipment operation and maintenance device of the embodiment of the present application, by real-time collection of the operation and maintenance data of the train control equipment, and entity recognition processing, and extracting entity knowledge information, the transformation from static analysis to dynamic monitoring is realized, and through real-time data collection, abnormal conditions of the equipment can be discovered in time, and rapid responses can be made to avoid the accumulation of problems caused by data lags. By fusing the entity knowledge information in the operation and maintenance data, constructing the operation and maintenance knowledge graph of the train control equipment, and using the knowledge graph technology, the operating status of the equipment can be deeply analyzed, and multi-level and multi-dimensional monitoring of the operating status of the equipment can be realized. Then, the operation and maintenance knowledge graph is imported into the database of the digital twin model, and by simulating the equipment operating status under different environmental conditions, the performance of the train control equipment in extreme environments is predicted and evaluated. Based on these prediction results, reasonable operation and maintenance strategies can be formulated, thereby improving the reliability and safety of the train control equipment in complex environments.

[0141] In a possible implementation manner, the simulation prediction module 503 is specifically used to:

[0142] The digital twin model is used to establish the association between the operation and maintenance knowledge graph and the real-time status data of the train control equipment. Based on the association, the operating status of the train control equipment is simulated to obtain real-time operating data.

[0143] Based on the real-time operation data and the historical operation and maintenance data of the train control equipment, the future operation status of the train control equipment is predicted.

[0144] In a possible implementation manner, the simulation prediction module 503 is specifically used to:

[0145] Based on the future operating status of the virtual model of the train control equipment, the fault diagnosis information in the operation and maintenance knowledge graph is called;

[0146] Determine the fault type based on the fault diagnosis information;

[0147] Generate an operation and maintenance strategy for train control equipment based on the fault type.

[0148] In a possible implementation, the operation and maintenance data includes: operation logs, operation abnormality analysis reports, vehicle equipment failure information, and maintenance solutions; the entity processing module 502 is specifically used to:

[0149] Clean the operation and maintenance data to obtain processed operation and maintenance data;

[0150] Perform entity recognition on the processed operation and maintenance data to extract entity knowledge information.

[0151] In a possible implementation manner, the entity processing module 502 is specifically configured to:

[0152] Perform labeling on the processed operation and maintenance data to obtain labeled operation and maintenance data;

[0153] Input the annotated operation and maintenance data into the pre-trained deep learning model, and convert the annotated operation and maintenance data into an operation and maintenance vector based on the deep learning model;

[0154] Extracting contextual feature information from the operation and maintenance vector;

[0155] Determine the label sequence corresponding to the context feature information, and extract entity knowledge information based on the label sequence.

[0156] In a possible implementation manner, the entity processing module 502 is specifically configured to:

[0157] Build entity dictionary based on entity knowledge information;

[0158] For each entity in the entity dictionary, determine the similarity between the entity and any remaining entity in the entity dictionary except the entity;

[0159] Based on the similarity, determine whether to merge the entity with the remaining entities;

[0160] Based on the fused entities, an operation and maintenance knowledge graph is constructed.

[0161] In a possible implementation manner, the entity processing module 502 is specifically configured to:

[0162] The similarity is compared with a preset threshold. If the similarity is greater than the preset threshold, it is determined to merge the entity and the remaining entities into one entity, and to delete duplicate entities in the entity dictionary. The duplicate entity is any entity in the entity and the remaining entities.

[0163] In a possible implementation manner, the entity processing module 502 is further configured to:

[0164] The entity knowledge information is input into the pre-trained hybrid network model, and the relationship of the entity knowledge information is extracted based on the hybrid network model to determine the entity association relationship. The entity association relationship includes: the association relationship between the equipment name and the fault type, and the association relationship between the fault type and the maintenance plan.

[0165] In a possible implementation manner, the entity processing module 502 is specifically configured to:

[0166] Based on the fused entities and entity associations, an operation and maintenance knowledge graph is constructed.

[0167] For descriptions of the processing flow of each module in the device and the interaction flow between each module, reference may be made to the relevant descriptions in the above method embodiment, which will not be described in detail here.

[0168] The present application embodiment also provides an electronic device 600, such as Figure 6 As shown, it is a schematic diagram of the structure of an electronic device 600 provided in an embodiment of the present application, including: a processor 601, a memory 602, and optionally, a bus 603. The memory 602 stores machine-readable instructions executable by the processor 601. When the electronic device 600 is running, the processor 601 communicates with the memory 602 through the bus 603. When the machine-readable instructions are executed by the processor 601, the steps of the train control device operation and maintenance method as described in any one of the above items are performed.

[0169] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the train control equipment operation and maintenance method as described in any one of the above items are executed.

[0170] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the system and device described above can refer to the corresponding process in the method embodiment, and will not be repeated in this application. In the several embodiments provided in this application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0171] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or part of the technical solution that contributes to the prior art or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Me mory), disk or optical disk and other media that can store program code.

[0172] The above are only specific implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be covered by the protection scope of the present application.

Claims

1. A train control equipment operation and maintenance method, characterized in that: include: Initialize the digital twin model of the train control device based on the device information of the train control device; Collecting operation and maintenance data of the train control equipment in real time, performing entity recognition processing on the operation and maintenance data, extracting entity knowledge information from the operation and maintenance data, and fusing the entity knowledge information to construct an operation and maintenance knowledge graph of the train control equipment, wherein the entity knowledge information includes: equipment name, fault type, and maintenance plan; The operation and maintenance knowledge graph is imported into the database of the digital twin model, the actual operation status of the train control device is simulated through the digital twin model, the future operation status of the train control device is predicted, and the operation and maintenance strategy of the train control device is determined based on the future operation status of the virtual model of the train control device.

2. The method according to claim 1, characterized in that The simulating the actual operating state of the train control device by the digital twin model and predicting the future operating state of the train control device includes: Establishing an association relationship between the operation and maintenance knowledge graph and the real-time status data of the train control device through the digital twin model, simulating the operating status of the train control device based on the association relationship, and obtaining real-time operating data; Based on the real-time operation data and the historical operation and maintenance data of the train control device, the future operation status of the train control device is predicted.

3. The method according to claim 1, characterized in that The determining the operation and maintenance strategy of the train control device based on the future operation state of the train control device virtual model includes: Based on the future operating state of the virtual model of the train control device, calling the fault diagnosis information in the operation and maintenance knowledge graph; Determining the fault type according to the fault diagnosis information; Based on the fault type, an operation and maintenance strategy of the train control device is generated.

4. The method according to claim 1, characterized in that: The operation and maintenance data include: operation logs, operation abnormality analysis reports, vehicle equipment failure information, and maintenance and disposal plans; After collecting the operation and maintenance data of the train control device, the method further includes: Performing data cleaning on the operation and maintenance data to obtain processed operation and maintenance data; The performing entity recognition processing on the operation and maintenance data to extract entity knowledge information in the operation and maintenance data includes: Perform entity recognition processing on the processed operation and maintenance data to extract the entity knowledge information.

5. The method according to claim 4, characterized in that The performing entity recognition processing on the processed operation and maintenance data to extract the entity knowledge information includes: Performing labeling processing on the processed operation and maintenance data to obtain labeled operation and maintenance data; Inputting the annotated operation and maintenance data into a pre-trained deep learning model, and converting the annotated operation and maintenance data into an operation and maintenance vector based on the deep learning model; Extracting context feature information from the operation and maintenance vector; A tag sequence corresponding to the context feature information is determined, and based on the tag sequence, the entity knowledge information is extracted.

6. The method according to claim 1, characterized in that The fusing and processing the entity knowledge information to construct the operation and maintenance knowledge graph of the train control device includes: Building an entity dictionary based on the entity knowledge information; For each entity in the entity dictionary, determining a similarity between the entity and any remaining entity in the entity dictionary except the entity; Based on the similarity, determining whether to perform fusion processing on the entity and the remaining entities; Based on the fused entities, the operation and maintenance knowledge graph is constructed.

7. The method according to claim 6, characterized in that The determining whether to perform fusion processing on the entity and the remaining entities based on the similarity includes: The similarity is compared with a preset threshold. If the similarity is greater than the preset threshold, it is determined to fuse the entity and the remaining entities into one entity, and to delete duplicate entities in the entity dictionary. The duplicate entity is any entity between the entity and the remaining entities.

8. The method according to claim 1, characterized in that Before fusing the entity knowledge information to construct the operation and maintenance knowledge graph of the train control device, the method further includes: The entity knowledge information is input into a pre-trained hybrid network model, and relationship extraction is performed on the entity knowledge information based on the hybrid network model to determine entity association relationships, wherein the entity association relationships include: an association relationship between a device name and a fault type, and an association relationship between a fault type and a maintenance plan.

9. The method according to claim 8, characterized in that The step of constructing the operation and maintenance knowledge graph based on the fused entities includes: Based on the fused entities and the entity association relationships, the operation and maintenance knowledge graph is constructed.

10. A train control equipment operation and maintenance device, characterized in that: include: An initialization module, used to initialize the digital twin model of the train control device based on the device information of the train control device; An entity processing module is used to collect the operation and maintenance data of the train control equipment in real time, perform entity recognition processing on the operation and maintenance data, extract entity knowledge information from the operation and maintenance data, and perform fusion processing on the entity knowledge information to construct an operation and maintenance knowledge graph of the train control equipment, wherein the entity knowledge information includes: equipment name, fault type, and maintenance plan; A simulation prediction module is used to import the operation and maintenance knowledge graph into the database of the digital twin model, simulate the actual operation status of the train control equipment through the digital twin model, predict the future operation status of the train control equipment, and determine the operation and maintenance strategy of the train control equipment based on the future operation status of the virtual model of the train control equipment.

Images