Intelligent train operation control method and system based on artificial intelligence

Abstract

The invention provides an intelligent train operation control method and system based on artificial intelligence, and relates to the technical field of artificial intelligence. Train state sensing data are partitioned, and graph coding vector data of a plurality of train state sensing paths are generated; key features are effectively extracted from complex train state sensing data through the feature focusing process, accurate target state sensing features are generated, and the recognition precision of the train running state is further improved. Moreover, the method achieves the dynamic and real-time optimization of the operation state of the train through the determination of the to-be-optimized state cooperation chain, and guarantees the stability and safety of the operation of the train. And finally, intelligent control is performed on the target train based on the to-be-optimized state coordination chain, so that the train operation efficiency is improved, and the risk caused by manual operation is reduced.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an artificial intelligence-based train operation intelligent control method and system. Background Art

[0002] As an important part of the modern transportation system, the operating efficiency and safety of trains have always been the focus of the industry. Traditional train control methods are mainly based on manual operations and fixed rule systems. This control method can play a good role when the train operation state is relatively simple and stable. However, with the increasing complexity of the train operation environment and the continuous improvement of operation requirements, traditional control methods have gradually exposed limitations such as delayed response and insufficient control accuracy.

[0003] In order to improve the accuracy and real-time performance of train operation control, the industry has been exploring new technical means. Among them, making full use of train status perception data has become an important research direction. During the operation of the train, a large amount of status data will be generated, including speed, position, acceleration, braking force and other information. These data contain the real-time status and future trends of the train operation, which is of great significance for optimizing the train control strategy.

[0004] However, existing train status data processing methods are often relatively simple and difficult to fully and accurately extract key information from these data. Some methods rely solely on a single train status data, ignoring the correlation and complementarity between different status data; other methods consider multiple status data but lack effective data processing and analysis methods, resulting in an inability to accurately reflect the actual situation of train operation. Summary of the invention

[0005] In view of the above-mentioned problems, in combination with the first aspect of the present invention, an embodiment of the present invention provides an intelligent control method for train operation based on artificial intelligence, the method comprising: taking multiple train state perception data in the train state perception data sequence of the target train as candidate train state perception data respectively, dividing the candidate train state perception data into blocks to generate multiple train state perception blocks, each of the train state perception blocks including multiple train state perception paths; Performing graph encoding on the multiple train state perception paths respectively to generate graph encoding vector data of the multiple train state perception paths; The multiple train state perception blocks are respectively used as candidate blocks, feature focusing is performed on the graph coded vector data of multiple train state perception paths in the candidate blocks to generate feature focused data of the candidate blocks, and the feature focused data of the multiple train state perception blocks are fused to generate target state perception features of the candidate train state perception data; Feature focusing is performed on target state perception features of multiple train state perception data in the train state perception data sequence to generate feature focused data of the train state perception data sequence, and a state coordination chain to be optimized of the train state perception data sequence is determined according to the feature focused data of the train state perception data sequence; The target train is intelligently controlled based on the coordinated chain of states to be optimized of the train state sensing data sequence.

[0006] In a possible implementation of the first aspect, the taking the multiple train state perception blocks as candidate blocks respectively, performing feature focusing on the graph coded vector data of the multiple train state perception paths in the candidate blocks, and generating feature focused data of the candidate blocks includes: The plurality of train state perception blocks are respectively used as candidate blocks, and feature focusing is performed on the graph encoding vector data of the plurality of train state perception paths in the candidate blocks to generate first feature focusing data, wherein the first feature focusing data includes a plurality of first focusing clusters; Determine the feature focusing data of the candidate block based on the feature focusing data of the plurality of first focusing clusters; The step of performing feature focusing on target state perception features of a plurality of train state perception data in the train state perception data sequence to generate feature focused data of the train state perception data sequence includes: Performing feature focusing on target state perception features of a plurality of train state perception data in the train state perception data sequence to generate second feature focused data, wherein the second feature focused data includes a plurality of second focused subgroups; The feature focused data of the train state perception data sequence is determined based on the feature focused data of the plurality of second focused clusters.

[0007] In a possible implementation of the first aspect, determining the feature focused data of the candidate block based on the feature focused data of the multiple first focused clusters includes: Fusing the representation vectors of the cluster centers of the multiple first focused clusters to generate feature focused data of the candidate blocks; The determining the feature focused data of the train state perception data sequence based on the feature focused data of the plurality of second focused subgroups comprises: The characterization vectors of the cluster centers of the multiple second focused clusters are fused to generate feature focused data of the train state perception data sequence.

[0008] In a possible implementation of the first aspect, the taking the multiple train state perception blocks as candidate blocks respectively, and performing feature focusing on the graph coded vector data of the multiple train state perception paths in the candidate blocks to generate first feature focused data includes: The plurality of train state perception blocks are respectively used as candidate blocks, and according to a plurality of different first clustering quantities, the graph encoding vector data of the plurality of train state perception paths in the candidate blocks are feature focused to generate first feature focused data associated with each of the plurality of first clustering quantities, wherein the first clustering quantity represents the size of the first focused cluster in the corresponding first feature focused data; the feature focused data of the candidate blocks include the feature focused data associated with each of the plurality of first clustering quantities of the candidate blocks; The step of fusing the feature focused data of the plurality of train state perception blocks to generate the target state perception features of the candidate train state perception data includes: Fusion of feature focused data corresponding to the same first cluster quantity in the feature focused data of the plurality of train state perception blocks to generate target state perception features of the candidate train state perception data corresponding to a plurality of first cluster quantities respectively; The step of performing feature focusing on target state perception features of a plurality of train state perception data in the train state perception data sequence to generate second feature focused data includes: Among the target state perception features of multiple train state perception data in the train state perception data sequence, feature focusing is performed on the target state perception features corresponding to the same first cluster quantity to generate second feature focused data corresponding to multiple first cluster quantities respectively; the feature focused data of the train state perception data sequence includes feature focused data associated with each of the multiple first cluster quantities of the train state perception data sequence.

[0009] In a possible implementation manner of the first aspect, each second feature focused data includes feature focused sub-data associated with N different second cluster quantities, each feature focused sub-data includes a plurality of second focused sub-clusters; the second cluster quantity represents the size of the second focused sub-clusters in the corresponding second feature focused data; The feature focused data of each first cluster number of the train state perception data sequence includes a plurality of sub-vector data of the first cluster number of the train state perception data sequence respectively corresponding to a plurality of second cluster numbers.

[0010] In a possible implementation of the first aspect, the step of determining the state coordination chain to be optimized of the train state perception data sequence according to the feature focused data of the train state perception data sequence includes: Performing state-space mapping on the feature-focused data of the train state perception data sequence to generate state-space mapping vector data of the train state perception data sequence; A decision is made on the state space mapping vector data of the train state perception data sequence to generate a state coordination chain to be optimized of the train state perception data sequence.

[0011] In a possible implementation of the first aspect, the making a decision on the state space mapping vector data of the train state perception data sequence to generate a state coordination chain to be optimized of the train state perception data sequence includes: Based on the train control knowledge expressed by the train control knowledge base, an artificial intelligence network model is used to make decisions on the state space mapping vector data of the train state perception data sequence, and determine the state coordination chain to be optimized of the train state perception data sequence.

[0012] In a possible implementation of the first aspect, the train control knowledge expressed by the train control knowledge base is used to make a decision on the state space mapping vector data of the train state perception data sequence using an artificial intelligence network model to determine the state coordination chain to be optimized of the train state perception data sequence, including: fusing the state space mapping vector data of the train state perception data sequence and the collaborative guidance vector data of the train state perception data sequence to generate global vector data of the train state perception data sequence; Based on the train control knowledge expressed by the train control knowledge base, using an artificial intelligence network model, making decisions on the global vector data of the train state perception data sequence, and determining a state coordination chain to be optimized of the train state perception data sequence; Among them, the step of making a decision on the global vector data of the train state perception data sequence based on the train control knowledge expressed by the train control knowledge base and determining the state coordination chain to be optimized of the train state perception data sequence by using an artificial intelligence network model includes: Collect and organize various professional knowledge resources in the field of train operation, and use natural language processing technology and knowledge engineering technology to extract a structured train control knowledge base from the various professional knowledge resources. The train control knowledge base includes train operation rules, safety restrictions, equipment operation logic, and fault handling procedures; Representing each knowledge feature in the extracted train control knowledge base in the form of nodes and edges, generating a corresponding knowledge network, and extracting a key feature set in the global vector data using a feature selection algorithm, wherein the key feature set reflects a key change trend of the train running state; Traversing each key feature in the key feature set, using the relationship network in the knowledge network, searching for association information of other entities and attributes associated with the key feature, and introducing context information for the original features in the knowledge network according to the association information to generate an enhanced feature set, wherein the context information includes the status of related components, historical fault records, and the impact of control instructions; The enhanced feature set is input into a pre-trained artificial intelligence network model to output a predicted train state coordination chain, wherein the pre-trained artificial intelligence network model incorporates a decision rule set pre-developed based on the train control knowledge base.

[0013] In a possible implementation manner of the first aspect, performing state space mapping on the feature focused data of the train state perception data sequence to generate state space mapping vector data of the train state perception data sequence includes: Parsing the structure of the feature-focused data, determining the core data component and the auxiliary data component in the feature-focused data, and initializing relevant parameters for state-space mapping based on the core data component and the auxiliary data component, wherein the relevant parameters include the dimension setting of the mapping space and the initial weight of the mapping algorithm; For each core data component and auxiliary data component in the feature focused data, feature dimension conversion is performed according to a preset conversion rule to generate feature focused data after feature dimension conversion, wherein the conversion rule is defined based on the physical characteristics of the train running state and the inherent logical relationship between the train running state data; Construct a state space framework for train state mapping. During the construction process, determine the meaning and value range of the coordinate axis of the state space, the coordinate axis is used to reflect the various core state factors of the train state, and the key operating parameters of the train-related equipment are used as auxiliary coordinate axes, and define the boundaries and characteristics of different state areas in the state space framework; Each core data component and auxiliary data component in the feature focused data is mapped to the constructed state space framework according to the corresponding coordinate axes. During the mapping process, the positions of the core data components and the auxiliary data components in the state space framework are determined according to the numerical values ​​of the core data components and the auxiliary data components and the definitions of the coordinate axes in the state space framework. After the mapping is completed, the mapping positions of each core data component and the auxiliary data component in the state space framework are combined to generate the state space mapping vector.

[0014] On the other hand, an embodiment of the present invention also provides an artificial intelligence-based train operation intelligent control system, including a processor and a machine-readable storage medium, wherein the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above method.

[0015] Based on the above aspects, the embodiment of the present application divides the train state perception data into blocks and generates graph-encoded vector data of multiple train state perception paths. Secondly, the key features are effectively extracted from the complex train state perception data through the feature focusing process, and accurate target state perception features are generated, thereby further improving the recognition accuracy of the train operation status. Furthermore, the present invention realizes dynamic and real-time optimization of the train operation status by determining the collaborative chain of states to be optimized, thereby ensuring the stability and safety of the train operation. Finally, the intelligent control of the target train based on the collaborative chain of states to be optimized not only improves the efficiency of the train operation, but also reduces the risks brought by human operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the execution flow of the train operation intelligent control method based on artificial intelligence provided by an embodiment of the present invention.

[0017] Figure 2 It is a schematic diagram of the hardware architecture of the artificial intelligence-based train operation intelligent control system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The present invention will be described in detail below with reference to the accompanying drawings. Figure 1 It is a flow chart of an intelligent control method for train operation based on artificial intelligence provided by an embodiment of the present invention. The intelligent control method for train operation based on artificial intelligence is introduced in detail below.

[0019] Step S110, taking multiple train state perception data in the train state perception data sequence of the target train as candidate train state perception data, dividing the candidate train state perception data into blocks to generate multiple train state perception blocks, each of the train state perception blocks including multiple train state perception paths.

[0020] In this embodiment, the target train is equipped with a large number of sensors, which continuously collect various status information during the operation of the train, and these information constitute the train status perception data sequence. For example, the sensor can collect the train's speed, acceleration, temperature and humidity in the carriage, working state parameters of various key equipment (such as braking system, traction system, etc.) (such as braking pressure, current of traction motor, etc.), and the train's position information on the track.

[0021] After receiving these massive train state perception data sequences, the server starts to process them. Taking the train state perception data collected in one time period as an example, the server uses each data in the train state perception data sequence in this time period as candidate train state perception data. For example, the train speed data at a certain moment is taken as a candidate train state perception data, and the temperature data in the carriage at the same moment is taken as another candidate train state perception data, etc.

[0022] Then, the server performs a block operation on each candidate train state perception data. Assume that the server performs block operations according to different subsystems or different functional areas of the train. Taking the traction system of the train as an example, the server can divide all state perception data related to the traction system (such as the current, voltage, temperature, etc. of the traction motor) into a train state perception block; for the braking system, the relevant data such as the braking pressure and the degree of wear of the brake pads are divided into another train state perception block. Each such train state perception block contains multiple train state perception paths. For example, in the train state perception block of the traction system, the process from the power supply of the traction motor to the operation of the motor, and then to the conversion of energy into the traction force of the train can be regarded as a train state perception path; and the process from the issuance of the control signal to the current adjustment of the traction motor can be regarded as another train state perception path. These different paths together constitute multiple train state perception paths in the train state perception block of the train traction system. Similarly, in the train state perception block of the braking system, the process from issuing braking commands to adjusting braking pressure and then to feedback on the actual braking effect of the train is a train state perception path. The process from monitoring brake pad wear to adjusting braking strategy according to the degree of wear is another train state perception path, and so on.

[0023] Step S120: Graph encoding is performed on the multiple train state perception paths respectively to generate graph encoding vector data of the multiple train state perception paths.

[0024] In this embodiment, a train state perception path in the traction system of the train is taken as an example. This train state perception path is from the power supply of the traction motor to the operation of the motor, and then to the conversion of energy into the traction force of the train. The server first determines the various nodes in this train state perception path, such as the power supply node, the winding node inside the motor (because the operation of the motor is related to the electromagnetic induction of the winding), the coupling node of energy transmission, and the friction node between the train wheel and the track (because the traction force ultimately acts between the wheel and the track). Then, the server constructs a graph structure based on the relationship between these nodes. For example, there is a power transmission relationship between the power supply node and the winding node inside the motor, then a directed edge is used in the graph to represent this relationship, and the weight of the edge can be determined according to factors such as the efficiency or power of the power transmission; there is a magnetic field coupling relationship between the winding node inside the motor and the coupling node of the energy transmission, which is also represented by a directed edge and the corresponding weight is determined; there is a force transmission relationship between the coupling node of the energy transmission and the friction node between the train wheel and the track, and the corresponding directed edge and weight are also constructed.

[0025] After the graph structure is constructed, the server uses a specific graph coding algorithm to encode and represent the graph structure. Assuming that a graph coding algorithm based on an adjacency matrix is ​​used, the server converts the relationship between nodes and edges in the graph into a graph coding vector data in the form of a matrix. In this adjacency matrix, the rows and columns of the matrix correspond to the nodes in the graph respectively. If there is an edge connection between two nodes, the weight of the edge is recorded at the corresponding position of the matrix; if there is no connection, it is recorded as 0. In this way, the corresponding graph coding vector data is generated for this train state perception path. In the same way, the server performs graph coding on other train state perception paths in the traction system and all train state perception paths in other train state perception blocks such as the braking system and the car environment control system, and generates their own graph coding vector data.

[0026] Step S130, respectively taking the multiple train state perception blocks as candidate blocks, performing feature focusing on the graph encoding vector data of the multiple train state perception paths in the candidate blocks, generating feature focusing data of the candidate blocks, and fusing the feature focusing data of the multiple train state perception blocks to generate the target state perception features of the candidate train state perception data.

[0027] In this embodiment, the multiple train state perception blocks generated previously can be used as candidate blocks respectively, and a feature focusing operation can be performed on the graph encoded vector data of the multiple train state perception paths in each candidate block.

[0028] Take the train's traction system as an example, which contains multiple train state perception paths for which graph-encoded vector data have been generated before. The server first clusters these graph-encoded vector data according to a certain feature focusing algorithm, such as a density-based spatial clustering algorithm, to generate first feature focusing data. In this process, the server clusters similar graph-encoded vector data together to form multiple first focusing clusters based on the feature spatial distribution of the graph-encoded vector data. For example, graph-encoded vector data related to the normal operating state of the traction motor (such as graph-encoded vector data related to normal power supply, normal winding electromagnetic induction, etc.) may be clustered into a first focusing cluster, while graph-encoded vector data related to the special operating state of the traction motor under high load (such as graph-encoded vector data related to high current, high magnetic field strength, etc.) may be clustered into another first focusing cluster.

[0029] Then, the server determines the feature focus data of this candidate block (traction system) based on the feature focus data of these multiple first focus clusters. Specifically, the server regards each first focus cluster as a whole and calculates the cluster center of each first focus cluster. For example, for a first focus cluster containing multiple graph encoding vector data, the representation vector of the cluster center is determined by calculating the mean vector of these vector data. Finally, the server merges the representation vectors of the cluster centers of these first focus clusters to generate the feature focus data of the traction system candidate block.

[0030] For other candidate blocks (such as the braking system, the car environment control system, etc.), the server also uses the above method to obtain the feature focus data of each candidate block. Then, the server fuses the feature focus data of multiple train state perception blocks (such as the traction system, the braking system, the car environment control system, etc.). For example, the server performs a weighted summation of the corresponding elements in the feature focus data of different candidate blocks according to a certain weight distribution rule to generate the target state perception features of the candidate train state perception data. Assuming that the weight of the feature focus data of the traction system in the target state perception features is 0.4, the weight of the braking system is 0.3, and the weight of the car environment control system is 0.3, then the feature focus data of each candidate block is fused according to this weight.

[0031] Step S140, feature focusing the target state perception features of multiple train state perception data in the train state perception data sequence to generate feature focused data of the train state perception data sequence, and determining the state collaborative chain to be optimized of the train state perception data sequence based on the feature focused data of the train state perception data sequence.

[0032] In this embodiment, after obtaining the target state perception features of each candidate train state perception data in the train state perception data sequence, a feature focusing operation is performed on these target state perception features to generate feature focused data of the train state perception data sequence.

[0033] Taking multiple candidate train state perception data (such as target state perception features such as speed, brake pressure, and car temperature) of a train within a period of time as an example, the server uses a clustering algorithm similar to the previous step S130 to cluster these target state perception features and generate second feature focus data. In this process, multiple second focus clusters will be formed. For example, those target state perception features related to the safe operation of the train (such as target state perception features related to normal brake pressure, reasonable speed range, etc.) may be clustered into a second focus cluster, while those target state perception features related to the comfort of train passengers (such as target state perception features related to suitable car temperature, humidity, etc.) may be clustered into another second focus cluster.

[0034] Then, the server determines the feature focus data of the train state perception data sequence based on the feature focus data of the multiple second focus clusters. The specific operation is to fuse the representation vectors of the cluster center of each second focus cluster together, for example, by weighted summation.

[0035] After obtaining the feature-focused data of the train state perception data sequence, the server begins to determine the state coordination chain to be optimized of the train state perception data sequence based on this feature-focused data. First, the server performs a state space mapping operation on the feature-focused data of the train state perception data sequence. For example, the server parses the structure of the feature-focused data and determines the core data components (such as the key operating parameters of the train, such as speed, brake pressure, etc.) and auxiliary data components (such as environmental parameters in the car, etc.). Then, based on these core data components and auxiliary data components, the server initializes the relevant parameters for state space mapping. For example, if the core data component is the train speed, the server determines the speed dimension setting of the mapping space according to the design speed range of the train; if the auxiliary data component is the car temperature, its relevant parameters in the state space mapping are determined according to the normal temperature range.

[0036] Next, for each core data component and auxiliary data component in the feature focused data, the feature dimension conversion is performed according to the pre-set conversion rules. For example, for the core data component of train speed, if the original data is the actual speed value (such as 100 kilometers per hour), it is converted according to the pre-set conversion rules (such as mapping the speed value to a value between 0-1, assuming that 100 kilometers per hour corresponds to 0.5). Then, the server builds a state space framework for train state mapping. In this framework, the meaning of the coordinate axes of the state space is determined, such as taking the train speed as one coordinate axis, the brake pressure as another coordinate axis, and the car temperature as the third coordinate axis, and the value range of each coordinate axis is determined. At the same time, the key operating parameters of the train-related equipment are used as auxiliary coordinate axes, and the boundaries and characteristics of different state areas are defined in this state space framework. For example, the area corresponding to the speed, brake pressure and car temperature range of the normal operation of the train is defined as the normal operation area, and the area corresponding to the speed, brake pressure and other parameters close to the safety limit is defined as the warning area, etc.

[0037] Finally, the server maps each core data component and auxiliary data component in the feature-focused data to the constructed state space framework according to the corresponding coordinate axis to generate state space mapping vector data. Then, based on the train control knowledge expressed by the train control knowledge base, the artificial intelligence network model is used to make decisions on this state space mapping vector data. For example, the server obtains knowledge such as the train operation rules (such as speed limits on different sections, braking distance requirements under different load conditions, etc.), safety restrictions (such as maximum braking pressure, maximum operating speed, etc.), equipment operation logic (such as the collaborative operation logic of the traction system and the braking system), and fault handling procedures (such as emergency handling procedures when the braking system fails) from the train control knowledge base. This knowledge is combined with the state space mapping vector data, and a pre-trained artificial intelligence network model (such as a deep neural network model) is used to determine the state coordination chain to be optimized for the train state perception data sequence after a series of calculations and reasoning. For example, the artificial intelligence network model determines how the train's traction system and braking system should work together in the next period of time based on the current train state (represented by state space mapping vector data) and train control knowledge (such as appropriately reducing the speed when entering a curve, the traction system reducing power output, and the braking system being ready, etc.). This is the state collaborative chain to be optimized.

[0038] Step S150: Intelligently control the target train based on the coordinated chain of states to be optimized of the train state sensing data sequence.

[0039] In this embodiment, for example, a train is about to enter a speed-limited section, and the coordinated chain of the state to be optimized indicates that the train needs to reduce its speed before entering the section. The server first sends a control instruction to the traction system of the train, calculates the power output that the traction system needs to reduce based on the current speed of the train and the distance to the speed-limited section, and then sends this precise power adjustment instruction to the controller of the traction system, so that the traction motor gradually reduces its power output, thereby causing the train speed to begin to decrease.

[0040] At the same time, the server will also coordinate control with the train's braking system. Although the speed is mainly reduced by the traction system, the braking system also needs to be prepared in case the speed drops too slowly or unexpected situations occur. Based on the information in the collaborative chain of the state to be optimized, the server sends a pre-instruction to the braking system to inform the braking system of the current speed adjustment of the train and the possible need for braking assistance. For example, if the train speed is still higher than the speed limit at a certain distance from the speed limit section, the server will send more specific braking instructions to the braking system, such as instructions to increase the brake pressure, so that the braking system starts working to ensure that the train can reach the specified speed limit when entering the speed limit section.

[0041] During the entire process, the server will continue to monitor the train's state perception data sequence, such as continuously obtaining data such as the train's real-time speed, brake pressure, and car temperature. If it is found that the actual operating state of the train deviates from the expected state in the state collaborative chain to be optimized, such as the train speed drops too fast or too slow, the server will recalculate and adjust the control instructions based on the new state perception data. For example, if the train speed drops too fast, the server will reduce the brake pressure of the brake system or increase the power output of the traction system to maintain the smooth operation of the train and ensure that the train runs safely and efficiently in accordance with the requirements of the state collaborative chain to be optimized.

[0042] Based on the above steps, the embodiment of the present application divides the train state perception data into blocks and generates graph-encoded vector data of multiple train state perception paths. Secondly, the key features are effectively extracted from the complex train state perception data through the feature focusing process, and accurate target state perception features are generated, thereby further improving the recognition accuracy of the train operation status. Furthermore, the present invention realizes dynamic and real-time optimization of the train operation status by determining the collaborative chain of states to be optimized, thereby ensuring the stability and safety of the train operation. Finally, the intelligent control of the target train based on the collaborative chain of states to be optimized not only improves the efficiency of the train operation, but also reduces the risks brought by human operation.

[0043] In a possible implementation, step S130 includes: Step S131, taking the multiple train state perception blocks as candidate blocks respectively, and performing feature focusing on the graph encoding vector data of the multiple train state perception paths in the candidate blocks to generate first feature focusing data, wherein the first feature focusing data includes multiple first focusing clusters.

[0044] Step S132: determining the feature focusing data of the candidate blocks based on the feature focusing data of the multiple first focusing clusters.

[0045] In this embodiment, work is first carried out on each candidate block. For example, the traction system of the train is used as a train state perception block, and the train state perception path in this block covers the paths corresponding to the processes from power supply to motor operation, and then to traction force generation. The graph-encoded vector data of these paths contain information about the relationship between each node and related parameters. The server performs feature focusing on these graph-encoded vector data and uses a clustering algorithm to generate first feature-focused data. In this process, similar data are clustered together to form multiple first focused clusters based on the distribution characteristics of the graph-encoded vector data in the feature space. For example, graph-encoded vector data related to the normal and stable operation of the motor, such as normal voltage, current supply, and stable magnetic field generation, form a first focused cluster; and graph-encoded vector data related to the motor under special working conditions, such as high current and high magnetic field changes during overload, form another first focused cluster.

[0046] After obtaining these first focus clusters, the server determines the feature focus data of the candidate blocks based on their feature focus data. For each first focus cluster, the server calculates its cluster center. Taking a first focus cluster containing multiple graph-encoded vector data as an example, a characterization vector of the cluster center is obtained by performing specific mathematical calculations on these vector data, such as finding the mean vector. Then, the characterization vectors of the cluster centers of the various first focus clusters are fused, for example, by weighted summation, and the fused result is the feature focus data of the candidate block (such as the traction system).

[0047] Step S140 includes: Step S141, feature focusing is performed on target state perception features of multiple train state perception data in the train state perception data sequence to generate second feature focused data, where the second feature focused data includes multiple second focused clusters.

[0048] Step S142: determining the feature focusing data of the train state perception data sequence based on the feature focusing data of the multiple second focusing clusters.

[0049] In this embodiment, when the target state perception features of multiple train state perception data in the train state perception data sequence are feature focused to generate feature focused data of the train state perception data sequence, the operation process is similar. Taking the target state perception features of train speed, brake pressure, car temperature, etc. in the train state perception data sequence as an example, the server uses a clustering algorithm to feature focus these target state perception features to generate second feature focused data. In this process, multiple second focused clusters are formed based on the feature space distribution of the target state perception features. For example, target state perception features related to train safety, such as a reasonable brake pressure range, a speed value within a normal speed limit, etc., form a second focused cluster; and target state perception features related to train passenger comfort, such as a suitable car temperature, etc., form another second focused cluster.

[0050] Afterwards, the server determines the feature focus data of the train state perception data sequence based on the feature focus data of multiple second focus clusters. For each second focus cluster, the characterization vector of its cluster center is also calculated, for example, by calculating the mean of the target state perception characteristics within the focus cluster. Then, the characterization vectors of the cluster centers of these second focus clusters are fused. The fusion method can be weighted summation according to specific weights, and the final result is the feature focus data of the train state perception data sequence. This data lays the foundation for the subsequent determination of the state coordination chain to be optimized of the train state perception data sequence. The server can further analyze the running status of the train based on this data and make corresponding control decisions.

[0051] In a possible implementation, step S132 includes: The characterization vectors of the cluster centers of the multiple first focused clusters are fused to generate feature focused data of the candidate blocks.

[0052] Step S142 includes: The characterization vectors of the cluster centers of the multiple second focused clusters are fused to generate feature focused data of the train state perception data sequence.

[0053] In a possible implementation, step S131 includes: The plurality of train state perception blocks are respectively used as candidate blocks, and based on a plurality of different first clustering quantities, feature focusing is performed on the graph encoding vector data of a plurality of train state perception paths in the candidate blocks, to generate first feature focusing data associated with each of the plurality of first clustering quantities, wherein the first clustering quantity represents the size of the first focusing cluster in the corresponding first feature focusing data. The feature focusing data of the candidate blocks includes the feature focusing data associated with each of the plurality of first clustering quantities of the candidate blocks.

[0054] Step S130 may further include: Step S133, merging the feature focused data corresponding to the same first cluster quantity in the feature focused data of the multiple train state perception blocks to generate target state perception features of the candidate train state perception data corresponding to multiple first cluster quantities respectively.

[0055] Step S141 may include: performing feature focusing on the target state perception features corresponding to the same first cluster quantity among the target state perception features of the multiple train state perception data in the train state perception data sequence, and generating second feature focusing data corresponding to the multiple first cluster quantities respectively. The feature focusing data of the train state perception data sequence includes feature focusing data associated with each of the multiple first cluster quantities of the train state perception data sequence.

[0056] In a possible implementation, each second feature focused data includes feature focused sub-data associated with N different second cluster quantities, and each feature focused sub-data includes multiple second focused sub-clusters. The second cluster quantity characterizes the size of the second focused sub-clusters in the corresponding second feature focused data. The feature focused data of each first cluster quantity of the train state perception data sequence includes multiple sub-vector data of the first cluster quantity of the train state perception data sequence corresponding to multiple second cluster quantities.

[0057] In this embodiment, the traction system of the train is taken as an example of the train state perception block. In this traction system block, the first feature focus data containing multiple first focus clusters has been generated by feature focusing on the graph encoding vector data of multiple train state perception paths. Each first focus cluster has a characterization vector of its cluster center, which is obtained by performing specific calculations on the graph encoding vector data within the focus cluster. For example, for a first focus cluster containing several graph encoding vector data, the mean vector of these vector data may be calculated as the characterization vector of the cluster center.

[0058] The server fuses the characterization vectors of the cluster centers of each first focus cluster. Assume that there are three first focus clusters in the candidate block of the traction system, namely, the first focus cluster related to the normal operation of the motor, the first focus cluster related to the motor overload protection, and the first focus cluster related to the motor starting process. For the first focus cluster related to the normal operation of the motor, the characterization vector of its cluster center contains the mean information of graph-encoded vector data features such as voltage, current, and magnetic field strength during normal operation; the cluster center characterization vector of the first focus cluster related to the motor overload protection contains the mean information of graph-encoded vector data features such as special current and temperature during overload; the cluster center characterization vector of the first focus cluster related to the motor starting process contains the mean information of graph-encoded vector data features such as voltage change and current rise during startup. The server fuses the characterization vectors of the three cluster centers according to certain rules, such as weighted summation according to pre-set weights. If the weight of the first focus cluster related to normal operation is set to 0.5, the weight related to overload protection is set to 0.3, and the weight related to the startup process is set to 0.2, then the characterization vectors of the centers of each cluster are weighted and summed according to this weight, and the result is the feature focus data of the candidate block of the traction system. This feature focus data integrates the key feature information of the traction system under different operating conditions and can comprehensively reflect the overall status of the traction system.

[0059] Similarly, when the server determines the feature focus data of the train state perception data sequence based on the feature focus data of multiple second focus clusters, the target state perception features such as train speed, brake pressure, and car temperature in the train state perception data sequence are taken as an example. These target state perception features have formed multiple second focus clusters through the previous clustering operation. Each second focus cluster also has a characterization vector of its cluster center, which is obtained by calculating the target state perception features within the focus cluster. For example, for the second focus cluster related to train safety, the characterization vector of its cluster center may be obtained by calculating the mean of target state perception features such as a reasonable brake pressure range and a speed value within the normal speed limit within the focus cluster; for the second focus cluster related to train passenger comfort, the characterization vector of its cluster center may be obtained by calculating the mean of target state perception features such as a suitable car temperature.

[0060] The server fuses the characterization vectors of the cluster centers of these second focus clusters. Assuming there are two second focus clusters, the weight of the second focus cluster related to safety assurance is set to 0.6, and the weight related to passenger comfort is set to 0.4. The characterization vectors of the cluster centers are weighted and summed according to this weight, and the result is the feature focus data of the train state perception data sequence. This data integrates key information from different aspects of train operation, which is of great significance for the subsequent analysis of the overall state of the train and control decision-making.

[0061] When the server uses multiple train state perception blocks as candidate blocks, and performs feature focusing on the graph coded vector data of multiple train state perception paths in the candidate blocks according to multiple different first cluster numbers to generate first feature focused data associated with each of the multiple first cluster numbers, the train state perception block of the traction system of the train is taken as an example. Assume that three different first cluster numbers are set, namely 5, 8 and 10.

[0062] When the number of first clusters is 5, the server clusters the graph coded vector data of multiple train state perception paths in the traction system according to a certain clustering algorithm, so that the first feature focused data with a scale of 5 first focused sub-clusters is finally formed. In this process, the server will cluster the data together according to the similarity and other features between the graph coded vector data to form these 5 first focused sub-clusters. The graph coded vector data within each first focused sub-clusters have high similarity in certain features, for example, they may be clustered together with the graph coded vector data related to a specific sub-function or a specific working condition in the traction system.

[0063] When the number of first clusters is 8 and 10, the server also performs clustering operations on the graph coding vector data according to the respective clustering requirements, and obtains the first feature focusing data of the first focusing subgroups of 8 and 10 respectively. In this way, the feature focusing data of the candidate block of the traction system includes the feature focusing data associated with each of the three different first cluster numbers (5, 8 and 10). The first feature focusing data under each different number of clusters reflects the status information of the traction system from different clustering granularities.

[0064] When the server fuses the feature focus data of multiple train state perception blocks to generate the target state perception features of candidate train state perception data, it is assumed that in addition to the traction system train state perception block, there are two train state perception blocks, the braking system and the car environment control system. Each block has its own feature focus data corresponding to a different number of first clusters (such as 5, 8, and 10).

[0065] For the case where the first cluster number is 5, the server fuses the feature focus data of the three blocks of traction system, braking system and cabin environment control system when the first cluster number is 5. For example, the feature focus data of the traction system when the first cluster number is 5 contains some key operating status information of the traction motor, the feature focus data of the braking system under this cluster number contains key information related to braking such as braking pressure and brake pad wear, and the feature focus data of the cabin environment control system under this cluster number contains environmental information such as cabin temperature and humidity. The server fuses the feature focus data of the three blocks when the first cluster number is 5 according to certain fusion rules, such as weighted summation or other specific algorithms, to generate the target state perception features of the candidate train state perception data when the first cluster number is 5.

[0066] Similarly, for the cases where the number of first clusters is 8 and 10, the feature focus data corresponding to the three blocks are fused in the same manner to generate the target state perception features of the candidate train state perception data when the number of first clusters is 8 and 10, respectively.

[0067] When the server performs feature focusing on the target state perception features of multiple train state perception data in the train state perception data sequence to generate second feature focused data, the target state perception features such as train speed, brake pressure, and car temperature in the train state perception data sequence are taken as an example, and the target state perception features corresponding to different first clustering numbers (5, 8, and 10) generated previously are considered.

[0068] For the case where the number of first clusters is 5, the server performs feature focusing on the target state perception features corresponding to the first cluster number of 5 in the target state perception features such as train speed, brake pressure, and car temperature. Assume that each target state perception feature contains rich information, for example, the target state perception feature of train speed contains speed information under different sections and different loads. The server performs clustering operations on these target state perception features corresponding to the first cluster number of 5 through a clustering algorithm to generate second feature focusing data corresponding to the first cluster number of 5. This second feature focusing data contains multiple feature focusing sub-data related to different second cluster numbers, and each feature focusing sub-data contains multiple second focusing sub-clusters. For example, there may be a feature focusing sub-data with a second cluster number of 3 related to key parameters for train operation safety, in which the three second focusing sub-clusters focus on information such as the safety threshold of train speed, the safety range of brake pressure, and the reasonable range of car temperature.

[0069] Similarly, for the cases where the number of first clusters is 8 and 10, similar feature focusing operations are performed on the corresponding target state perception features to generate second feature focusing data corresponding to the first cluster numbers of 8 and 10. The feature focusing data of each first cluster number (5, 8 and 10) of the train state perception data sequence contains multiple sub-vector data corresponding to multiple second cluster numbers. These data reflect the feature information of the train state perception data sequence from different levels and angles, providing a rich data basis for the subsequent determination of the state coordination chain to be optimized of the train state perception data sequence.

[0070] In a possible implementation, step S140 further includes: Step S143, performing state-space mapping on the feature-focused data of the train state perception data sequence to generate state-space mapping vector data of the train state perception data sequence.

[0071] Step S144, making a decision on the state space mapping vector data of the train state perception data sequence to generate a state coordination chain to be optimized for the train state perception data sequence.

[0072] In a possible implementation, step S144 includes: based on the train control knowledge expressed by the train control knowledge base, using an artificial intelligence network model to make decisions on the state space mapping vector data of the train state perception data sequence, and determining the state coordination chain to be optimized of the train state perception data sequence.

[0073] In this embodiment, the feature focus data of the previously obtained train state perception data sequence is taken as an example, which contains key information integrated from various subsystems of the train (such as traction system, braking system, carriage environment control system, etc.). The server begins to parse the structure of the feature focus data and determine the core data components and auxiliary data components. The core data components may be key parameters directly related to the operation of the train, such as the speed of the train, the braking pressure, the power output of the traction system, etc.; the auxiliary data components may be data such as the temperature and humidity in the carriage, which do not directly affect the operation of the train but have a certain correlation with the overall state. Based on these core data components and auxiliary data components, the server initializes the relevant parameters for state space mapping. For example, for the core data component of train speed, if the design speed range of the train is 0-200 kilometers per hour, then in the state space mapping, the value range of the speed dimension is set to this interval; for the auxiliary data component of carriage temperature, if the normal carriage temperature range is 18-26 degrees Celsius, the relevant parameter range in the state space mapping is set.

[0074] Next, for each core data component and auxiliary data component in the feature-focused data, the feature dimension is converted according to the pre-set conversion rules. Taking the train speed as an example, if the original speed data is 120 km / h, it may be converted to 0.6 according to the conversion rules (assuming that the speed value is mapped to a value between 0-1, the corresponding ratio of 120 km / h in the range of 0-200 km / h is 0.6). Then the server builds a state space framework for train state mapping. In this framework, the meaning of the coordinate axis of the state space is determined, with the train speed as one coordinate axis, the brake pressure as another coordinate axis, and the car temperature as the third coordinate axis. At the same time, the value range of each coordinate axis is determined, such as the speed coordinate axis value range is 0-1 (the range after conversion), and the brake pressure coordinate axis sets the corresponding value range according to the pressure range of the brake system. In addition, the key operating parameters of train-related equipment are used as auxiliary coordinate axes to define the boundaries and characteristics of different state areas in this state space framework. For example, the area corresponding to the value range of speed, brake pressure and car temperature when the train is operating normally is defined as the normal operating area, and the area corresponding to the value range when the speed is close to the maximum value, the brake pressure is at a high value and the car temperature is abnormal is defined as the danger warning area, etc.

[0075] Finally, the server maps each core data component and auxiliary data component in the feature-focused data to the constructed state space frame according to the corresponding coordinate axes. During the mapping process, the positions of the core data components and auxiliary data components in the state space frame are determined according to the values ​​of the core data components and auxiliary data components and the definitions of the coordinate axes in the state space frame. After the mapping is completed, the mapping positions of each core data component and auxiliary data component in the state space frame are combined to generate state space mapping vector data.

[0076] After obtaining the state space mapping vector data of the train state perception data sequence, the server needs to make a decision to generate the state coordination chain to be optimized for the train state perception data sequence. This process is based on the train control knowledge expressed by the train control knowledge base and is implemented using an artificial intelligence network model.

[0077] The server first obtains relevant train control knowledge from the train control knowledge base, which contains information such as train operation rules, safety restrictions, equipment operation logic, and fault handling procedures. For example, the train operation rules stipulate speed limits on different sections, and the speed must be reduced to a certain value on curves; the safety restrictions specify the maximum braking pressure of the braking system, the maximum operating speed of the train, etc.; the equipment operation logic describes how the traction system and the braking system work together, such as reducing power output through the traction system when decelerating, and then starting the braking system according to the situation; the fault handling process includes emergency operation procedures when the braking system fails, etc.

[0078] Then, the server combines the state space mapping vector data of the train state perception data sequence with these train control knowledge and inputs them into the pre-trained artificial intelligence network model. This artificial intelligence network model may be a deep neural network model that has been trained with a large amount of train operation data and corresponding control strategy data. Based on the input data, the artificial intelligence network model determines the state coordination chain to be optimized for the train state perception data sequence after a series of calculations and reasoning. For example, based on the current train speed, brake pressure, and car temperature and other state information (represented by state space mapping vector data), combined with the train control knowledge such as the speed limit and safety requirements of the current section, the model determines how the train's traction system and braking system should work together in the next period of time. If the train is approaching a speed limit section, the model may determine that the traction system should gradually reduce power output, the braking system is ready, and when it is a certain distance away from the speed limit section, if the speed is still higher than the speed limit, the braking system begins to apply appropriate braking pressure to ensure that the train can reach the specified speed limit when entering the speed limit section. This is the determined state coordination chain to be optimized.

[0079] In a possible implementation, the train control knowledge expressed by the train control knowledge base is used to make a decision on the state space mapping vector data of the train state perception data sequence using an artificial intelligence network model to determine the state coordination chain to be optimized of the train state perception data sequence, including: Step S1441, fusing the state space mapping vector data of the train state perception data sequence and the collaborative guidance vector data of the train state perception data sequence to generate global vector data of the train state perception data sequence.

[0080] Step S1442, based on the train control knowledge expressed by the train control knowledge base, using the artificial intelligence network model, make decisions on the global vector data of the train state perception data sequence, and determine the state coordination chain to be optimized of the train state perception data sequence.

[0081] Wherein, step S1442 includes: Step S1442-1, collect and organize various professional knowledge resources in the field of train operation, and use natural language processing technology and knowledge engineering technology to extract a structured train control knowledge base from the various professional knowledge resources. The train control knowledge base includes train operation rules, safety restrictions, equipment operation logic, and fault handling procedures.

[0082] Step S1442-2, each knowledge feature extracted from the train control knowledge base is represented in the form of nodes and edges. After generating the corresponding knowledge network, a feature selection algorithm is used to extract a key feature set from the global vector data, and the key feature set reflects the key change trend of the train operation status.

[0083] Step S1442-3, traverse each key feature in the key feature set, use the relationship network in the knowledge network to find the association information of other entities and attributes associated with the key feature, and introduce context information for the original features in the knowledge network based on the association information to generate an enhanced feature set, wherein the context information includes the status of related components, historical fault records, and the impact of control instructions.

[0084] Step S1442-4, inputting the enhanced feature set into a pre-trained artificial intelligence network model, and outputting a predicted train state collaborative chain, wherein the pre-trained artificial intelligence network model incorporates a decision rule set pre-developed based on the train control knowledge base.

[0085] In this embodiment, the state space mapping vector data of the train state perception data sequence obtained previously is taken as an example, which contains the state information of various aspects of the train after state space mapping, such as the mapping values ​​of speed, brake pressure, car temperature, etc. in the state space. The collaborative guidance vector data of the train state perception data sequence may contain some specific information for guiding the coordination of train operation, such as guidance information related to the distance between the front and rear trains in the scenario of multi-train collaborative operation, or priority information for collaborative operation between different subsystems of the train. The server fuses the two vector data according to a specific fusion algorithm. For example, a weighted summation method can be used to set weights according to the importance of the state space mapping vector data and the collaborative guidance vector data in reflecting the overall state of the train. If the state space mapping vector data is more important in reflecting the current state of the train, the weight may be set to 0.7, and the weight of the collaborative guidance vector data in reflecting the collaborative relationship between the train and the outside is set to 0.3. According to this weight, the corresponding elements of the two vector data are added to generate the global vector data of the train state perception data sequence. This global vector data integrates information on both the train's own state and the collaborative relationship, providing a more comprehensive data basis for subsequent decision-making.

[0086] Next, the server uses an artificial intelligence network model to make decisions on the global vector data of the train state perception data sequence based on the train control knowledge expressed by the train control knowledge base to determine the state coordination chain to be optimized for the train state perception data sequence.

[0087] The first step is to collect and organize various professional knowledge resources in the field of train operation, and use natural language processing technology and knowledge engineering technology to extract a structured train control knowledge base from these resources. The professional knowledge resources in the field of train operation are very rich, including various technical documents, operating manuals, safety specifications, etc. formulated by the railway department. For example, in terms of train operation rules, there are speed limit regulations for different lines, such as the speed limit of 80 kilometers per hour at certain curves on mountain lines and the speed limit of 160 kilometers per hour on plain straight sections; there are also operation priority rules for different types of trains (such as passenger trains and freight trains). In terms of safety restrictions, the maximum braking pressure of the train braking system is clarified, for example, the maximum braking pressure of a certain model of train braking system is 1000 kPa, and the maximum axle weight limit of the train. The equipment operation logic includes operating procedures such as how the traction system adjusts the power output according to the command and how the braking system accurately controls the braking pressure according to the braking command. The fault handling process covers how to switch to the backup equipment if the traction motor of the train fails, and how to take emergency braking measures when the braking system fails.

[0088] The server uses natural language processing technology to parse these documents, identify key information, and then structure this information through knowledge engineering technology to build a train control knowledge base. For example, speed limit rules are classified and stored according to line type, train type, etc.; the safety limit parameters of the braking system are indexed according to the equipment model, etc.

[0089] After constructing the train control knowledge base, each knowledge feature is represented in the form of nodes and edges to generate a corresponding knowledge network. Taking the speed limit in the train operation rules as an example, the speed limit values ​​of different lines are used as nodes, and the corresponding relationship between line types and speed limits is used as edges; for the equipment operation logic, the different operating states of the traction system are used as nodes, and the conversion conditions between the operating states are used as edges. Then, the key feature set in the global vector data is extracted using the feature selection algorithm. This feature selection algorithm may be based on principles such as information gain or correlation analysis. For example, in the global vector data, data such as train speed, brake pressure, and the distance between the front and rear trains have a great influence on the trend of the train's operating state. These features that reflect the key trend of the train's operating state are screened out through the feature selection algorithm to form a key feature set.

[0090] Next, the server traverses each key feature in the key feature set and uses the relationship network in the knowledge network to find the associated information of other entities and attributes associated with the key feature. Taking the key feature of train speed as an example, the server searches for entities and attributes related to it in the knowledge network. It may be found that entities and attributes associated with train speed include the slope of the line where the train is located (because the slope affects the control of the train speed) and the load of the current train (different loads have different speed control strategies). Based on this associated information, context information is introduced for the original features in the knowledge network to generate an enhanced feature set. For example, for the original feature of train speed, in addition to the speed value itself, context information such as the current line slope and train load is also added. If the train speed is 120 km / h, the line slope is 5%, and the load is 500 tons, then this information is combined to form a richer enhanced feature.

[0091] Finally, the enhanced feature set is input into the pre-trained artificial intelligence network model, and the predicted train state coordination chain is output. This pre-trained artificial intelligence network model incorporates a decision rule set that is pre-formulated based on the train control knowledge base. When training this artificial intelligence network model, a large amount of historical train operation data is used, which includes train state information, control instructions, and actual operation results. At the same time, a decision rule set is formulated according to the rules in the train control knowledge base, such as the braking distance requirements at a specific speed and the traction power limit at different loads, and these rules are integrated into the model training process. After the enhanced feature set is input, the model predicts the future operation state of the train and determines the train state coordination chain based on the existing training results and the integrated decision rules. For example, based on the current train speed, load, line slope, and the distance between the front and rear trains, combined with the speed limit, safety requirements and other knowledge in the train control knowledge base, the model predicts how the train's traction system should adjust the power output, how the braking system should cooperate, and how the distance between the front and rear trains should be maintained in the next period of time. This train state coordination chain can guide the intelligent control of the train and ensure the safe and efficient operation of the train.

[0092] In a possible implementation, step S143 includes: Step S1431, analyze the structure of the feature-focused data, determine the core data components and auxiliary data components in the feature-focused data, and initialize relevant parameters for state space mapping based on the core data components and auxiliary data components, wherein the relevant parameters include the dimension setting of the mapping space and the initial weight of the mapping algorithm.

[0093] Step S1432, for each core data component and auxiliary data component in the feature-focused data, feature dimension conversion is performed according to a pre-set conversion rule to generate feature-focused data after feature dimension conversion, wherein the conversion rule is defined based on the physical characteristics of the train running state and the inherent logical relationship between the train running state data.

[0094] Step S1433, construct a state space framework for train state mapping. During the construction process, determine the meaning and value range of the coordinate axis of the state space. The coordinate axis is used to reflect the various core state factors of the train state, and the key operating parameters of the train-related equipment are used as auxiliary coordinate axes. The boundaries and characteristics of different state areas are defined in the state space framework.

[0095] Step S1434, map each core data component and auxiliary data component in the feature focused data to the constructed state space framework according to the corresponding coordinate axes. During the mapping process, determine the positions of the core data components and the auxiliary data components in the state space framework according to the numerical values ​​of the core data components and the auxiliary data components and the definitions of the coordinate axes in the state space framework. After the mapping is completed, combine the mapping positions of each core data component and auxiliary data component in the state space framework to generate the state space mapping vector.

[0096] In this embodiment, taking the train operation data as an example, the feature focus data is the comprehensive data obtained after a series of previous processing. The core data components may include key parameters that directly affect the train operation status, such as the speed of the train, the brake pressure of the brake system, and the power output of the traction system. For example, the train speed is an important representation of the train operation status, and its numerical value is directly related to the train's travel progress, safety, and spacing control with other trains; the brake pressure is a key indicator of the effectiveness of the brake system, which directly determines the train's deceleration ability; the power output of the traction system affects the train's acceleration and climbing ability. Auxiliary data components may be data such as the temperature and humidity in the carriage, which do not directly determine the train operation but have a certain correlation with the overall state of the train. For example, too high a temperature in the carriage may affect passenger comfort and indirectly affect the train's operating efficiency.

[0097] Based on these core data components and auxiliary data components, the server initializes the relevant parameters for state space mapping. In terms of the dimension setting of the mapping space, if the core data components are train speed, brake pressure and traction power output, then a three-dimensional mapping space is set. For train speed, the value range of the corresponding dimension in the mapping space is determined according to the design speed range of the train, such as 0-300 km / h; the brake pressure is determined according to the design pressure range of the brake system, such as 0-1000 kPa; the traction power output is determined according to the power range of the traction system, such as 0-5000 kW. At the same time, the initial weight of the mapping algorithm is set according to the relative importance of each core data component in reflecting the overall state of the train. Assuming that the train speed has the most critical impact on the train's operating state, the weight is set to 0.5, the brake pressure weight is 0.3, and the traction power output weight is 0.2.

[0098] Next, for each core data component and auxiliary data component in the feature focused data, the feature dimension is converted according to the pre-set conversion rules. These conversion rules are defined based on the physical characteristics of the train running state and the inherent logical relationship between the train running state data. Taking the train speed as an example, if the original train speed data is the actual speed value in kilometers per hour, such as 150 kilometers per hour, and in the state space mapping, it is hoped that the speed will be mapped to a numerical range between 0 and 1, according to the train speed design range of 0-300 kilometers per hour, 150 kilometers per hour can be converted to 0.5 (i.e. 150 / 300=0.5) through a simple linear conversion rule. For the brake pressure, if the original data is the actual pressure value in kilopascals, assuming it is 300 kilopascals, according to the mapping requirements of the brake pressure in the state space, according to its value range of 0-1000 kilopascals, it is converted to 0.3 (i.e. 300 / 1000=0.3). A similar conversion is used for the traction power output. Assuming the original power is 1500 kilowatts, it is converted to 0.3 (i.e., 1500 / 5000=0.3) based on the value range of 0-5000 kilowatts. For auxiliary data components such as cabin temperature, if the original temperature is 25 degrees Celsius, and the mapping range of cabin temperature in the state space is 18-28 degrees Celsius, it is converted to a value between 0-1 through the corresponding conversion rules. It is assumed that the converted value is 0.7 (the calculation method is based on the pre-set conversion rules related to the range of 18-28 degrees Celsius). In this way, feature-focused data after feature dimension conversion is generated.

[0099] Then, the server constructs a state space framework for train state mapping. During the construction process, the meaning and value range of the coordinate axes of the state space are determined. The core state factors of the train are used as coordinate axes, such as the train speed, brake pressure and traction power output mentioned above. The train speed coordinate axis means the speed of the train, and the value range is 0-1 (the range after conversion) determined previously; the brake pressure coordinate axis means the pressure applied by the brake system, and the value range is 0-1; the traction power output coordinate axis means the power provided by the traction system, and the value range is 0-1. At the same time, the key operating parameters of train-related equipment are used as auxiliary coordinate axes, such as the carriage temperature coordinate axis, which means the temperature condition in the carriage, and the value range is 0-1 after the previous conversion. And define the boundaries and characteristics of different state areas in this state space framework. For example, the area corresponding to the value range of speed, brake pressure, traction power output and carriage temperature during normal operation of the train is defined as the normal operation area. Assuming that the train speed in normal operation corresponds to a range of 0.3-0.7, the brake pressure range is 0.2-0.6, the traction power output range is 0.1-0.5, and the carriage temperature range is 0.4-0.8, then the area surrounded by these value ranges in this state space framework is the normal operation area. When the train speed is close to 1, the brake pressure is close to 1, and the traction power output is close to 0, it may be defined as an emergency braking area, which indicates that the train is undergoing a high-intensity braking operation.

[0100] Finally, the server maps each core data component and auxiliary data component in the feature focused data to the constructed state space framework according to the corresponding coordinate axis. During the mapping process, the position of the core data component and the auxiliary data component in the state space framework is determined according to the numerical value of the core data component and the auxiliary data component and the definition of the coordinate axis in the state space framework. For example, for the previously converted train speed value of 0.5, it is at the 0.5 position on the train speed coordinate axis; the brake pressure value is 0.3, it is at the 0.3 position on the brake pressure coordinate axis; the traction power output value is 0.3, it is at the 0.3 position on the traction power output coordinate axis; the car temperature value is 0.7, it is at the 0.7 position on the car temperature coordinate axis. After the mapping is completed, the mapping positions of each core data component and auxiliary data component in the state space framework are combined to generate a state space mapping vector. This state space mapping vector can be understood as a specific coordinate point of the current state of the train in this constructed state space framework, which integrates the state information of various aspects of the train, and provides an important data basis for subsequent decisions based on the train state, such as determining the state coordination chain to be optimized of the train state perception data sequence.

[0101] Figure 2The hardware structure of the train operation intelligent control system 100 based on artificial intelligence for implementing the above-mentioned train operation intelligent control method based on artificial intelligence provided by an embodiment of the present invention is shown as follows: Figure 2 As shown, the artificial intelligence-based train operation intelligent control system 100 may include a processor 110 , a machine-readable storage medium 120 , a bus 130 , and a communication unit 140 .

[0102] The machine-readable storage medium 120 may store data and / or instructions. In some embodiments, the machine-readable storage medium 120 may store data acquired from an external terminal. In some embodiments, the machine-readable storage medium 120 may store data and / or instructions that the artificial intelligence-based train operation intelligent control system 100 uses to execute or use to complete the exemplary method described in the present invention.

[0103] During the specific implementation process, one or more processors 110 execute computer executable instructions stored in the machine-readable storage medium 120, so that the processor 110 can execute the artificial intelligence-based intelligent control method for train operation in the above method embodiment. The processor 110, the machine-readable storage medium 120 and the communication unit 140 are connected through the bus 130, and the processor 110 can be used to control the sending and receiving actions of the communication unit 140.

[0104] The specific implementation process of the processor 110 can refer to the various method embodiments executed by the above-mentioned artificial intelligence-based train operation intelligent control system 100. The implementation principles and technical effects are similar, and this embodiment will not be repeated here.

[0105] In addition, an embodiment of the present invention further provides a readable storage medium, in which computer executable instructions are preset. When a processor executes the computer executable instructions, the above-mentioned intelligent control method for train operation based on artificial intelligence is implemented.

[0106] It should be noted that in order to simplify the description of the present invention and thus help understand one or more embodiments of the invention, in the foregoing description of the embodiments of the present invention, various features are sometimes combined into one embodiment, drawing or description thereof.

Claims

1. An intelligent train operation control method based on artificial intelligence, characterized in that: The method comprises: taking multiple train state perception data in the train state perception data sequence of the target train as candidate train state perception data respectively, dividing the candidate train state perception data into blocks to generate multiple train state perception blocks, each of the train state perception blocks including multiple train state perception paths; Performing graph encoding on the multiple train state perception paths respectively to generate graph encoding vector data of the multiple train state perception paths; The multiple train state perception blocks are respectively used as candidate blocks, feature focusing is performed on the graph coded vector data of multiple train state perception paths in the candidate blocks to generate feature focused data of the candidate blocks, and the feature focused data of the multiple train state perception blocks are fused to generate target state perception features of the candidate train state perception data; Feature focusing is performed on target state perception features of multiple train state perception data in the train state perception data sequence to generate feature focused data of the train state perception data sequence, and a state coordination chain to be optimized of the train state perception data sequence is determined according to the feature focused data of the train state perception data sequence; The target train is intelligently controlled based on the coordinated chain of states to be optimized of the train state sensing data sequence.

2. The train operation intelligent control method based on artificial intelligence according to claim 1 is characterized in that: The method of taking the plurality of train state perception blocks as candidate blocks respectively, performing feature focusing on the graph coded vector data of the plurality of train state perception paths in the candidate blocks, and generating feature focused data of the candidate blocks includes: The plurality of train state perception blocks are respectively used as candidate blocks, and feature focusing is performed on the graph encoding vector data of the plurality of train state perception paths in the candidate blocks to generate first feature focusing data, wherein the first feature focusing data includes a plurality of first focusing clusters; Determine the feature focusing data of the candidate block based on the feature focusing data of the plurality of first focusing clusters; The step of performing feature focusing on target state perception features of a plurality of train state perception data in the train state perception data sequence to generate feature focused data of the train state perception data sequence includes: Performing feature focusing on target state perception features of a plurality of train state perception data in the train state perception data sequence to generate second feature focused data, wherein the second feature focused data includes a plurality of second focused subgroups; The feature focused data of the train state perception data sequence is determined based on the feature focused data of the plurality of second focused clusters.

3. The train operation intelligent control method based on artificial intelligence according to claim 2 is characterized in that: The determining the feature focusing data of the candidate blocks based on the feature focusing data of the plurality of first focusing clusters comprises: Fusing the representation vectors of the cluster centers of the multiple first focused clusters to generate feature focused data of the candidate blocks; The determining the feature focused data of the train state perception data sequence based on the feature focused data of the plurality of second focused subgroups comprises: The characterization vectors of the cluster centers of the multiple second focused clusters are fused to generate feature focused data of the train state perception data sequence.

4. The train operation intelligent control method based on artificial intelligence according to claim 2 is characterized in that: The method of taking the plurality of train state perception blocks as candidate blocks respectively and performing feature focusing on the graph coded vector data of the plurality of train state perception paths in the candidate blocks to generate first feature focused data comprises: The plurality of train state perception blocks are respectively used as candidate blocks, and according to a plurality of different first clustering quantities, the graph encoding vector data of the plurality of train state perception paths in the candidate blocks are feature focused to generate first feature focused data associated with each of the plurality of first clustering quantities, wherein the first clustering quantity represents the size of the first focused cluster in the corresponding first feature focused data; the feature focused data of the candidate blocks include the feature focused data associated with each of the plurality of first clustering quantities of the candidate blocks; The step of fusing the feature focused data of the plurality of train state perception blocks to generate the target state perception features of the candidate train state perception data includes: Fusion of feature focused data corresponding to the same first cluster quantity in the feature focused data of the plurality of train state perception blocks to generate target state perception features of the candidate train state perception data corresponding to a plurality of first cluster quantities respectively; The step of performing feature focusing on target state perception features of a plurality of train state perception data in the train state perception data sequence to generate second feature focused data includes: Among the target state perception features of multiple train state perception data in the train state perception data sequence, feature focusing is performed on the target state perception features corresponding to the same first cluster quantity to generate second feature focused data corresponding to multiple first cluster quantities respectively; the feature focused data of the train state perception data sequence includes feature focused data associated with each of the multiple first cluster quantities of the train state perception data sequence.

5. The train operation intelligent control method based on artificial intelligence according to claim 3 is characterized in that: Each second feature focused data includes feature focused sub-data associated with N different second cluster numbers, each feature focused sub-data includes a plurality of second focused sub-clusters; the second cluster number represents the size of the second focused sub-clusters in the corresponding second feature focused data; The feature focused data of each first cluster number of the train state perception data sequence includes a plurality of sub-vector data of the first cluster number of the train state perception data sequence respectively corresponding to a plurality of second cluster numbers.

6. The train operation intelligent control method based on artificial intelligence according to any one of claims 1 to 5, characterized in that: The step of determining the state coordination chain to be optimized of the train state perception data sequence according to the feature focused data of the train state perception data sequence comprises: Performing state-space mapping on the feature-focused data of the train state perception data sequence to generate state-space mapping vector data of the train state perception data sequence; A decision is made on the state space mapping vector data of the train state perception data sequence to generate a state coordination chain to be optimized of the train state perception data sequence.

7. The train operation intelligent control method based on artificial intelligence according to claim 6 is characterized in that: The making a decision on the state space mapping vector data of the train state perception data sequence to generate a state coordination chain to be optimized of the train state perception data sequence includes: Based on the train control knowledge expressed by the train control knowledge base, an artificial intelligence network model is used to make decisions on the state space mapping vector data of the train state perception data sequence, and determine the state coordination chain to be optimized of the train state perception data sequence.

8. The train operation intelligent control method based on artificial intelligence according to claim 7 is characterized in that: The train control knowledge expressed by the train control knowledge base is used to make a decision on the state space mapping vector data of the train state perception data sequence by using an artificial intelligence network model to determine the state coordination chain to be optimized of the train state perception data sequence, including: fusing the state space mapping vector data of the train state perception data sequence and the collaborative guidance vector data of the train state perception data sequence to generate global vector data of the train state perception data sequence; Based on the train control knowledge expressed by the train control knowledge base, using an artificial intelligence network model, making decisions on the global vector data of the train state perception data sequence, and determining a state coordination chain to be optimized of the train state perception data sequence; Among them, the step of making a decision on the global vector data of the train state perception data sequence based on the train control knowledge expressed by the train control knowledge base and determining the state coordination chain to be optimized of the train state perception data sequence by using an artificial intelligence network model includes: Collect and organize various professional knowledge resources in the field of train operation, and use natural language processing technology and knowledge engineering technology to extract a structured train control knowledge base from the various professional knowledge resources. The train control knowledge base includes train operation rules, safety restrictions, equipment operation logic, and fault handling procedures; Representing each knowledge feature in the extracted train control knowledge base in the form of nodes and edges, generating a corresponding knowledge network, and extracting a key feature set in the global vector data using a feature selection algorithm, wherein the key feature set reflects a key change trend of the train running state; Traversing each key feature in the key feature set, using the relationship network in the knowledge network, searching for association information of other entities and attributes associated with the key feature, and introducing context information for the original features in the knowledge network according to the association information to generate an enhanced feature set, wherein the context information includes the status of related components, historical fault records, and the impact of control instructions; The enhanced feature set is input into a pre-trained artificial intelligence network model to output a predicted train state coordination chain, wherein the pre-trained artificial intelligence network model incorporates a decision rule set pre-developed based on the train control knowledge base.

9. The train operation intelligent control method based on artificial intelligence according to claim 6 is characterized in that: The performing state space mapping on the feature focused data of the train state perception data sequence to generate state space mapping vector data of the train state perception data sequence includes: Parsing the structure of the feature-focused data, determining the core data component and the auxiliary data component in the feature-focused data, and initializing relevant parameters for state-space mapping based on the core data component and the auxiliary data component, wherein the relevant parameters include the dimension setting of the mapping space and the initial weight of the mapping algorithm; For each core data component and auxiliary data component in the feature focused data, feature dimension conversion is performed according to a preset conversion rule to generate feature focused data after feature dimension conversion, wherein the conversion rule is defined based on the physical characteristics of the train running state and the inherent logical relationship between the train running state data; Constructing a state space framework for train state mapping, during which the meaning and value range of the coordinate axes of the state space are determined, the coordinate axes are used to reflect the core state factors of the train state, and the key operating parameters of the train-related equipment are used as auxiliary coordinate axes, and the boundaries and characteristics of different state areas are defined in the state space framework; Each core data component and auxiliary data component in the feature focused data is mapped to the constructed state space framework according to the corresponding coordinate axes. During the mapping process, the positions of the core data components and the auxiliary data components in the state space framework are determined according to the numerical values ​​of the core data components and the auxiliary data components and the definitions of the coordinate axes in the state space framework. After the mapping is completed, the mapping positions of each core data component and the auxiliary data component in the state space framework are combined to generate the state space mapping vector.

10. An intelligent train operation control system based on artificial intelligence, characterized in that: The artificial intelligence-based train operation intelligent control system includes a processor and a memory, the memory is connected to the processor, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to implement the artificial intelligence-based train operation intelligent control method described in any one of claims 1 to 9 above.

Images