A train operation electronic map construction method and system based on meta learning

Abstract

The application discloses a kind of based on meta learning's train operation electronic map construction method and system, the method first synchronously collects the geographic coordinate data of train high-precision positioning equipment RTK and the kilometre mark data of train operation monitoring device LKJ, time alignment and interpolation processing establish space-time alignment fusion sequence;Again, electronic map construction is converted into the regression problem of " longitude and latitude back kilometre mark", parameterized model is used to describe the nonlinear mapping relationship of two;Introduce deviation correction model to compensate RTK drift and kilometre mark error;Through meta learning mechanism, learn general initial parameters based on model-independent meta learning strategy in offline stage, and quickly adapt new line or dynamic deviation using a small number of samples in online stage;Finally, fusion correction data is generated and continuously updated electronic map.The application improves the precision of electronic map construction, new line adaptation capability and long-term stability, meets the high-precision positioning and real-time calibration requirements of rail transit.

Description

Technical Field

[0001] This invention relates to the field of rail transit and intelligent map construction technology, and in particular to a method and system for constructing electronic train operation maps based on meta-learning. Background Technology

[0002] Electronic maps play a crucial role in rail transit dispatching, command, and safety monitoring, and their accuracy directly affects the safety and efficiency of train operations. Existing methods for constructing train electronic maps often rely on single-source positioning information, which has significant limitations: for example, when relying solely on RTK positioning data, satellite signals are prone to interruption or drift and multipath errors in environments such as tunnels and building obstructions, leading to decreased positioning accuracy; when relying solely on kilometer marker data provided by LKJ (Train Operation Monitoring and Recording Equipment), this data only reflects the train's mileage progress on the physical line and cannot directly correspond to geographic coordinates, and there are measurement deviations and differences from the line reference points.

[0003] In existing technologies, non-point cloud data collected by vehicles is typically converted into point cloud format and compared with radar point clouds to verify the accuracy of electronic map data. While this method improves the automation and accuracy of map data verification in road scenarios, it does not address the mapping relationship between latitude and longitude and kilometer markers in the rail transit environment. In single-track scenarios with train operation, it cannot handle RTK positioning drift, kilometer marker errors, and dynamic deviations caused by environmental changes, and lacks adaptive modeling and online correction capabilities, making it difficult to meet the needs of high-precision electronic map construction and real-time positioning calibration for rail transit.

[0004] Furthermore, when new lines open or the operating conditions of existing lines change, traditional methods require a large amount of labeled data for model training to adapt, resulting in low adaptation efficiency, high costs, and an inability to achieve rapid deployment and application. Therefore, there is an urgent need for a method for constructing electronic train operation maps that can integrate multi-source positioning data, resist environmental interference, and quickly adapt to new lines. Summary of the Invention

[0005] This invention aims to solve at least one of the technical problems existing in the prior art, and proposes a method and system for constructing electronic maps of train operation based on meta-learning.

[0006] In a first aspect, embodiments of the present invention provide a method for constructing an electronic train operation map based on meta-learning, comprising:

[0007] Collect geographic coordinate data and mileage data, perform time alignment and interpolation processing on the geographic coordinate data and mileage data, and obtain a fusion sequence of kilometer marker data corresponding to geographic coordinate data under a unified time reference;

[0008] A parametric model is used to map and model the original kilometer marker data and geographic coordinate data. The kilometer marker data is then predicted and output based on the geographic coordinate data using the parametric model.

[0009] A deviation correction model is introduced, which takes the original kilometer marker data and input features as inputs and outputs the kilometer marker deviation to optimize the original kilometer marker data;

[0010] The loss function of the parameterized model is determined based on historical geographic coordinate data and mileage data. The parameterized model is then trained and optimized using meta-learning based on the loss function to obtain the optimized parameterized model.

[0011] The optimized kilometer marker data and geographic coordinate data are input into the optimized parametric model, and after smoothing, an electronic map in standard geographic information format is generated. New data generated by subsequent train operations are added as incremental samples to the training to achieve continuous optimization of the electronic map.

[0012] Furthermore, the geographic coordinate data output by the high-precision RTK positioning equipment for trains, including timestamps, latitude, and longitude, is represented as follows: ;in, For timestamps, , These are latitude and longitude coordinates, respectively. Geographic coordinate data reflects the actual running trajectory of the train in the geographic coordinate system.

[0013] Furthermore, the mileage data output by the train operation monitoring and recording equipment, which includes timestamps and kilometer markers, is represented as follows: ;in, For timestamps, This is the kilometer mark reading for the train; the mileage data represents the train's progress along the physical track.

[0014] Furthermore, the parameterized model expression is: ,in θ represents latitude and longitude, K represents the kilometer marker reading of the train, and θ is the model weight, which includes the weight matrix of each layer, the bias vector of each layer, and the learnable parameters related to the activation function. The learnable parameters are automatically learned by training the fusion sequence of kilometer marker data corresponding to the geographic coordinate data.

[0015] Furthermore, the expression for the deviation correction model is:

[0016] ;in This indicates the correction amount for the kilometer marker. Indicates the original kilometer marker. The input features include the number of visible satellites, position accuracy factor, horizontal accuracy factor, vertical accuracy factor parameterization, and train dynamics data ϕ parameterization, which form a parameter set for the deviation correction model. The deviation correction model is a small parameter neural network.

[0017] Furthermore, different operation processes of the same line are treated as different tasks, and a loss function is constructed. The loss function includes a regularization term to ensure that the mapping function is monotonically smooth; the expression of the loss function is:

[0018]

[0019] in, Representation of mapping model The overall loss function, This represents a parameterized kilometer marker prediction model. This represents the kilometer marker corresponding to the i-th sample. This represents the weighting parameter, which controls the degree of influence of the regularization term R. ; , Let represent the parameterized models of the (i+1)th and ith samples, respectively.

[0020] Furthermore, based on the loss function, a meta-learning method is used to train and optimize the parameterized model. Specific steps include:

[0021] During the offline model phase, multiple train operation data sets are acquired to form a task set. A model-independent meta-learning strategy is used to learn the model's general initial parameters, enabling the train to be matched with only a small number of samples when it is on a new line or in different positioning states. The expression for the general initial parameters is: ;in, The set of parameters representing the deviation correction model; This represents the "optimal initial parameters" obtained through meta-learning; This represents the validation loss function, used to evaluate whether the model generalizes well after one update. Indicates the first The training loss function for each task; Indicates the model parameters Find the gradient.

[0022] Furthermore, based on the loss function, a meta-learning method is used to train and optimize the parameterized model. Specific steps include:

[0023] During the online phase of the model, kilometer marker samples corresponding to geographical coordinates are acquired within a short time window at the beginning of train operation. Gradient updates are performed on the general initial parameters, and inference is conducted throughout the entire process based on the updated parameters. Periodic fine-tuning is then performed to train and optimize the parameterized model. The parameter update expression is: ; This indicates the corrected kilometer marker. This indicates the kilometer marker before correction.

[0024] Secondly, this invention also discloses a train operation electronic map construction system based on meta-learning, comprising: a data acquisition and preprocessing module, a parameterized model construction module, a raw kilometer marker data optimization module, a parameterized model optimization module, and an electronic map generation and management module; wherein:

[0025] The data acquisition and preprocessing module is used to acquire geographic coordinate data and mileage data, perform time alignment and interpolation on the geographic coordinate data and mileage data, and obtain a fusion sequence of kilometer marker data corresponding to geographic coordinate data under a unified time reference.

[0026] The parametric model building module is used to map and model the original kilometer marker data and geographic coordinate data using a parametric model, and to predict and output kilometer marker data based on geographic coordinate data using the parametric model.

[0027] The raw kilometer marker data optimization module is used to optimize the raw kilometer marker data by taking the raw kilometer marker data and input features as inputs and outputting the kilometer marker deviation through a deviation correction model.

[0028] The parameterized model optimization module is used to determine the loss function of the parameterized model based on historical geographic coordinate data and mileage data, and to train and optimize the parameterized model using meta-learning based on the loss function to obtain the optimized parameterized model.

[0029] The electronic map generation and management module is used to input optimized kilometer marker data and geographic coordinate data into the optimized parametric model, and generate an electronic map in a standard geographic information format after smoothing processing; new data generated by subsequent train operations are added as incremental samples to the training to achieve continuous optimization of the electronic map.

[0030] Thirdly, the present invention also discloses an electronic device, comprising:

[0031] One or more processors;

[0032] Memory, used to store one or more programs;

[0033] When the one or more programs are executed by the one or more processors, the one or more processors implement the method.

[0034] This invention discloses a method and system for constructing electronic train operation maps based on meta-learning, aiming to solve the problems of existing electronic train map construction relying on single positioning information, insufficient accuracy under occluded environments, and poor adaptability to new lines. The method first synchronously collects geographic coordinate data from the train high-precision positioning device RTK and kilometer marker data from the train operation monitoring device LKJ, and establishes a spatiotemporally aligned fusion sequence through time alignment and interpolation processing; then, the electronic map construction is transformed into a regression problem of "latitude and longitude back-derived kilometer markers," and a parametric model is used to characterize the nonlinear mapping relationship between the two; a deviation correction model is introduced to compensate for RTK drift and kilometer marker errors; through a meta-learning mechanism, in the offline stage, general initial parameters are learned based on a model-independent meta-learning strategy, and in the online stage, a small number of samples are used to quickly adapt to new lines or dynamic deviations; finally, the fused and corrected data are used to generate and continuously update the electronic map. This invention improves the accuracy of electronic map construction, the adaptability to new lines, and long-term stability, meeting the high-precision positioning and real-time calibration requirements of rail transit. Attached Figure Description

[0035] Figure 1 A flowchart illustrating a method for constructing an electronic train operation map based on meta-learning, provided in an embodiment of the present invention;

[0036] Figure 2 A structural block diagram of a train operation electronic map construction system based on meta-learning provided in an embodiment of the present invention;

[0037] Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0038] To enable those skilled in the art to better understand the technical solutions of the present invention, exemplary embodiments of the present invention are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0039] Where there is no conflict, the various embodiments of the present invention and the features thereof may be combined with each other.

[0040] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0041] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Terms such as “connected” or “linked” are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.

[0042] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant art and the invention, and will not be interpreted as having an idealized or overly formal meaning unless expressly so defined herein.

[0043] In the technical solution of this invention, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information all comply with relevant laws and regulations and do not violate public order and good morals. The use of user data in this technical solution follows relevant national laws and regulations (e.g., the "Information Security Technology - Personal Information Security Specification"). For example: appropriate measures are taken for personal information access control; restrictions are imposed on the display of personal information; the purpose of using personal information does not exceed the scope of direct or reasonable association; and explicit identity targeting is eliminated when using personal information to avoid precisely locating a specific individual.

[0044] To address at least one of the technical problems existing in the aforementioned related technologies, the present invention provides a method and system for constructing electronic maps of train operation based on meta-learning.

[0045] This embodiment discloses a method for constructing electronic train operation maps based on meta-learning, such as... Figure 1 ,include:

[0046] S100. Collect geographic coordinate data and mileage data, and perform time alignment and interpolation processing on the geographic coordinate data and mileage data to obtain a fused sequence of kilometer marker data corresponding to geographic coordinate data under a unified time reference; specifically, during train operation, data sources from two independent devices are collected simultaneously. By processing the two data sources, a spatiotemporally aligned data foundation is established to ensure that the subsequent learning model can associate the mapping relationship between kilometer markers and latitude and longitude on the same time scale:

[0047] In this embodiment, the geographic coordinate data output by the high-precision RTK positioning device for trains, including timestamps, latitude, and longitude, is represented as follows: ;in, For timestamps, , These are latitude and longitude coordinates, respectively. Geographic coordinate data reflects the actual running trajectory of the train in the geographic coordinate system.

[0048] In this embodiment, the mileage data output by the train operation monitoring and recording equipment, which includes timestamps and kilometer markers, is represented as follows: ;in, For timestamps, This is the kilometer mark reading for the train; the mileage data represents the train's progress along the physical track.

[0049] In this embodiment, since the two data sources have different sampling frequencies and time synchronization methods, time alignment and interpolation processing are required first to obtain a fused sequence under a unified time reference:

[0050]

[0051] This sequence ensures a one-to-one correspondence between geographical locations and kilometer marker data at every moment, providing an input basis for subsequent modeling.

[0052] S200. A parametric model is used to map and model the original kilometer marker data and geographic coordinate data. The kilometer marker data is predicted and output based on the geographic coordinate data using the parametric model. In this embodiment, the expression of the parametric model is: ,in θ represents latitude and longitude, K represents the kilometer marker reading of the train, and θ is the model weight, which includes the weight matrix of each layer, the bias vector of each layer, and the learnable parameters related to the activation function. The learnable parameters are automatically learned by training the fusion sequence of kilometer marker data corresponding to the geographic coordinate data.

[0053] Specifically, in this invention, the electronic map construction problem is transformed into a regression modeling problem of "deriving kilometer markers from latitude and longitude". Through a meta-learning mechanism, the mapping model is enabled to self-update and adapt to different operating environments, thereby improving the construction accuracy and long-term maintainability of the electronic map.

[0054] The core of electronic train operation maps is determining the functional relationship between kilometer markers and geographic coordinates:

[0055]

[0056] The mapping function It describes the latitude and longitude position of any mileage point on the train line and is the essential form of expression of electronic maps.

[0057] This function exhibits complex nonlinear characteristics due to factors such as the radius of the track curve, ranging error, satellite signal obstruction, and changes in train attitude, and the mapping relationship cannot be accurately described by a simple linear model. To accurately describe this relationship, this invention employs a parametric model:

[0058]

[0059] Where parameters The model weights are learnable parameters, including the weight matrices of each layer, the bias vectors of each layer, and the learnable parameters related to the activation functions. These parameters are automatically learned from a large amount of RTK-LKJ aligned geographic coordinate data during the training process, and determine the response of the neural network to output kilometer markers under different latitude and longitude inputs, reflecting factors such as the radius of the route curve and the distance measurement deviation.

[0060] S300. Introduce a deviation correction model, taking the original kilometer marker data and input features as input, outputting the kilometer marker deviation, and optimizing the original kilometer marker data; in this embodiment, the expression of the deviation correction model is: ;in This indicates the correction amount for the kilometer marker. Indicates the original kilometer marker. The input features include the number of visible satellites, position accuracy factor, horizontal accuracy factor, vertical accuracy factor parameterization, and train dynamics data ϕ parameterization, which form a parameter set for the deviation correction model. The deviation correction model is a small parameter neural network.

[0061] Specifically, in the process of electronic map generation, to further reduce the deviation caused by RTK drift or kilometer marker errors, this invention introduces a deviation correction model:

[0062]

[0063] in This indicates the corrected kilometer marker. Indicates the original kilometer marker. This indicates the correction amount for the kilometer marker. Represented as:

[0064]

[0065] Model Using the original kilometer marker and input features The input consists of features including the number of visible satellites, PDOP (Position Dilution of Precision), HDOP (Horizontal Dilution of Precision), VDOP (Vertical Dilution of Precision), and train dynamics. The output is a correction of the deviation, which is learned by establishing a small-parameter neural network. It has the advantages of fast computation, small memory usage, and the ability to be quickly fine-tuned with a small number of gradient steps.

[0066] S400. Determine the loss function of the parameterized model based on historical geographic coordinate data and mileage data, and use meta-learning method to train and optimize the parameterized model based on the loss function to obtain the optimized parameterized model;

[0067] In this embodiment, the mapping relationship between kilometer markers and geographical coordinates may slightly drift at different times or under different train operations on the same line. Therefore, this invention treats different operating processes on the same line as different tasks and collects historical calibration data. Specifically, different operation processes of the same line are treated as different tasks, and a loss function is constructed. This loss function includes a regularization term to ensure that the mapping function is monotonically smooth. The expression for the loss function is:

[0068]

[0069] in, Representation of mapping model The overall loss function, This represents a parameterized kilometer marker prediction model. This represents the kilometer marker corresponding to the i-th sample. This represents the weighting parameter, which controls the degree of influence of the regularization term R. ; , Let represent the parameterized models of the (i+1)th and ith samples, respectively.

[0070] In this embodiment, the parametric model is trained and optimized using a meta-learning method based on the loss function. Specific steps include:

[0071] During the offline model phase, multiple train operation data sets are acquired to form a task set. A model-independent meta-learning strategy is used to learn the model's general initial parameters, enabling the train to be matched with only a small number of samples when it is on a new line or in different positioning states. The expression for the general initial parameters is: ;in, The set of parameters representing the deviation correction model; This represents the "optimal initial parameters" obtained through meta-learning; This represents the validation loss function, used to evaluate whether the model generalizes well after one update. Indicates the first The training loss function for each task; Indicates the model parameters Find the gradient.

[0072] During the online phase of the model, kilometer marker samples corresponding to geographical coordinates are acquired within a short time window at the beginning of train operation. Gradient updates are performed on the general initial parameters, and inference is conducted throughout the entire process based on the updated parameters. Periodic fine-tuning is then performed to train and optimize the parameterized model. The parameter update expression is: ; This indicates the corrected kilometer marker. This indicates the kilometer marker before correction.

[0073] S500. The optimized kilometer marker data and geographic coordinate data are input into the optimized parametric model, and after smoothing, an electronic map in standard geographic information format is generated; new data generated by subsequent train operations are added as incremental samples to the training to achieve continuous optimization of the electronic map.

[0074] Specifically, an electronic map is generated by fusing the corrected coordinate data with the kilometer marker data. The generated map can be further smoothed and stored in a standard geographic information format. New data generated from each subsequent train run can be added as incremental samples to the training process, enabling continuous optimization of the map model and ensuring the long-term accuracy and adaptability of the electronic map.

[0075] This embodiment discloses a method and system for constructing electronic train operation maps based on meta-learning, aiming to solve the problems of existing electronic train map construction relying on single positioning information, insufficient accuracy under occluded environments, and poor adaptability to new lines. The method first synchronously collects geographic coordinate data from the train high-precision positioning device RTK and kilometer marker data from the train operation monitoring device LKJ, and establishes a spatiotemporally aligned fusion sequence through time alignment and interpolation processing; then, the electronic map construction is transformed into a regression problem of "latitude and longitude back-calculation of kilometer markers," and a parametric model is used to characterize the nonlinear mapping relationship between the two; a deviation correction model is introduced to compensate for RTK drift and kilometer marker errors; through a meta-learning mechanism, in the offline stage, general initial parameters are learned based on a model-independent meta-learning strategy, and in the online stage, a small number of samples are used to quickly adapt to new lines or dynamic deviations; finally, the fused correction data is used to generate and continuously update the electronic map. This invention improves the accuracy of electronic map construction, the adaptability to new lines, and long-term stability, meeting the high-precision positioning and real-time calibration requirements of rail transit.

[0076] Based on the same inventive concept, embodiments of the present invention also provide a train operation electronic map construction system based on meta-learning, such as... Figure 2 It includes: a data acquisition and preprocessing module, a parametric model construction module, a raw kilometer marker data optimization module, a parametric model optimization module, and an electronic map generation and management module; among which:

[0077] The data acquisition and preprocessing module is used to acquire geographic coordinate data and mileage data, perform time alignment and interpolation on the geographic coordinate data and mileage data, and obtain a fusion sequence of kilometer marker data corresponding to geographic coordinate data under a unified time reference.

[0078] The parametric model building module is used to map and model the original kilometer marker data and geographic coordinate data using a parametric model, and to predict and output kilometer marker data based on geographic coordinate data using the parametric model.

[0079] The raw kilometer marker data optimization module is used to optimize the raw kilometer marker data by taking the raw kilometer marker data and input features as inputs and outputting the kilometer marker deviation through a deviation correction model.

[0080] The parameterized model optimization module is used to determine the loss function of the parameterized model based on historical geographic coordinate data and mileage data, and to train and optimize the parameterized model using meta-learning based on the loss function to obtain the optimized parameterized model.

[0081] The electronic map generation and management module is used to input optimized kilometer marker data and geographic coordinate data into the optimized parametric model, and generate an electronic map in a standard geographic information format after smoothing processing; new data generated by subsequent train operations are added as incremental samples to the training to achieve continuous optimization of the electronic map.

[0082] Based on the same inventive concept, embodiments of the present invention also provide an electronic device. Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device including: one or more processors 101, a memory 102, and one or more I / O interfaces 103. The memory 102 stores one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement any of the methods described in the above embodiments; the one or more I / O interfaces 103 are connected between the processor and the memory, configured to enable information interaction between the processor and the memory.

[0083] The processor 101 is a device with data processing capabilities, including but not limited to a central processing unit (CPU); the memory 102 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and flash memory (FLASH); the I / O interface (read / write interface) 103 is connected between the processor 101 and the memory 102, and can realize information interaction between the processor 101 and the memory 102, including but not limited to a data bus (Bus).

[0084] In some embodiments, the processor 101, memory 102, and I / O interface 103 are interconnected via bus 104, and thus connected to other components of the computing device.

[0085] In some embodiments, the one or more processors 101 include a field-programmable gate array.

[0086] This invention also provides a computer-readable medium. The computer-readable medium stores a computer program, which, when executed by a processor, implements the steps of any of the methods described in the above embodiments. The computer-readable storage medium may be a volatile or non-volatile computer-readable storage medium.

[0087] This invention also provides a computer program product, including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code, wherein when the computer-readable code is run in the processor of an electronic device, the processor in the electronic device executes the above-described method.

[0088] Those skilled in the art will understand that all or some of the steps, systems, and apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software can be distributed on a computer-readable storage medium, which may include computer storage media (or non-transitory media) and communication media (or transient media).

[0089] As is known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable program instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), static random access memory (SRAM), flash memory or other memory technologies, portable compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, it is known to those skilled in the art that communication media typically contain computer-readable program instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0090] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0091] The computer program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the computer-readable program instructions. This electronic circuitry can execute the computer-readable program instructions to implement various aspects of the invention.

[0092] The computer program product described herein can be implemented specifically through hardware, software, or a combination thereof. In one alternative embodiment, the computer program product is specifically embodied in a computer storage medium; in another alternative embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0093] Various aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0094] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0095] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0096] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction, which contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0097] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A method for constructing an electronic train operation map based on meta-learning, characterized in that, include: Collect geographic coordinate data and mileage data, perform time alignment and interpolation processing on the geographic coordinate data and mileage data, and obtain a fusion sequence of kilometer marker data corresponding to geographic coordinate data under a unified time reference; A parametric model is used to map and model the original kilometer marker data and geographic coordinate data. The kilometer marker data is then predicted and output based on the geographic coordinate data using the parametric model. A deviation correction model is introduced, which takes the original kilometer marker data and input features as inputs and outputs the kilometer marker deviation to optimize the original kilometer marker data; The loss function of the parameterized model is determined based on historical geographic coordinate data and mileage data. The parameterized model is then trained and optimized using meta-learning based on the loss function to obtain the optimized parameterized model. The optimized kilometer marker data and geographic coordinate data are input into the optimized parametric model, and after smoothing, an electronic map in standard geographic information format is generated. New data generated during subsequent train operations are added as incremental samples to the training, enabling continuous optimization of the electronic map.

2. The method according to claim 1, characterized in that, The geographic coordinate data, including timestamp, latitude, and longitude, is output by the high-precision RTK positioning equipment for trains. This geographic coordinate data is represented as... ;in, For timestamps, , These are latitude and longitude coordinates, respectively. Geographic coordinate data reflects the actual running trajectory of the train in the geographic coordinate system.

3. The method according to claim 1, characterized in that, Mileage data output by the train operation monitoring and recording equipment, including timestamps and kilometer markers, is represented as follows: ;in, For timestamps, This is the kilometer mark reading for the train; the mileage data represents the train's progress along the physical track.

4. The method according to claim 1, characterized in that, The parameterized model expression is: ,in θ represents latitude and longitude, K represents the kilometer marker reading of the train, and θ is the model weight, which includes the weight matrix of each layer, the bias vector of each layer, and the learnable parameters related to the activation function. The learnable parameters are automatically learned by training the fusion sequence of kilometer marker data corresponding to the geographic coordinate data.

5. The method according to claim 1, characterized in that, The expression for the deviation correction model is: ;in This indicates the correction amount for the kilometer marker. Indicates the original kilometer marker. The input features include the number of visible satellites, position accuracy factor, horizontal accuracy factor, vertical accuracy factor parameterization, and train dynamics data ϕ parameterization, which form the parameter set of the deviation correction model. The deviation correction model is a small parameter neural network.

6. The method according to claim 1, characterized in that, Different operation processes of the same route are treated as different tasks, and a loss function is constructed. The loss function includes a regularization term to ensure that the mapping function is monotonically smooth; the expression of the loss function is: ； in, Representation of mapping model The overall loss function, This represents a parameterized kilometer marker prediction model. This represents the kilometer marker corresponding to the i-th sample. This represents the weighting parameter, which controls the degree of influence of the regularization term R. ; , Let represent the parameterized models of the (i+1)th and ith samples, respectively.

7. The method according to claim 1, characterized in that, Based on the aforementioned loss function, a meta-learning method is used to train and optimize the parameterized model. The specific steps include: During the offline model phase, multiple train operation data sets are acquired to form a task set. A model-independent meta-learning strategy is used to learn the model's general initial parameters, enabling the train to be matched with only a small number of samples when it is on a new line or in different positioning states. The expression for the general initial parameters is: ;in, The set of parameters representing the deviation correction model; This represents the "optimal initial parameters" obtained through meta-learning; This represents the validation loss function, used to evaluate whether the model generalizes well after one update. Indicates the first The training loss function for each task; Indicates the model parameters Find the gradient.

8. The method according to claim 7, characterized in that, Based on the aforementioned loss function, a meta-learning method is used to train and optimize the parameterized model. Specific steps also include: During the online phase of the model, kilometer marker samples corresponding to geographical coordinates are acquired within a short time window at the beginning of train operation. Gradient updates are performed on the general initial parameters, and inference is conducted throughout the entire process based on the updated parameters. Periodic fine-tuning is then performed to train and optimize the parameterized model. The parameter update expression is: ; This indicates the corrected kilometer marker. This indicates the kilometer marker before correction.

9. A train operation electronic map construction system based on meta-learning, employing the construction method described in any one of claims 1-8, characterized in that, include: The module comprises a data acquisition and preprocessing module, a parametric model construction module, a raw kilometer marker data optimization module, a parametric model optimization module, and an electronic map generation and management module; among which: The data acquisition and preprocessing module is used to acquire geographic coordinate data and mileage data, perform time alignment and interpolation on the geographic coordinate data and mileage data, and obtain a fusion sequence of kilometer marker data corresponding to geographic coordinate data under a unified time reference. The parametric model building module is used to map and model the original kilometer marker data and geographic coordinate data using a parametric model, and to predict and output kilometer marker data based on geographic coordinate data using the parametric model. The raw kilometer marker data optimization module is used to optimize the raw kilometer marker data by taking the raw kilometer marker data and input features as inputs and outputting the kilometer marker deviation through a deviation correction model. The parameterized model optimization module is used to determine the loss function of the parameterized model based on historical geographic coordinate data and mileage data, and to train and optimize the parameterized model using meta-learning based on the loss function to obtain the optimized parameterized model. The electronic map generation and management module is used to input optimized kilometer marker data and geographic coordinate data into the optimized parametric model, and generate an electronic map in a standard geographic information format after smoothing processing; new data generated by subsequent train operations are added as incremental samples to the training to achieve continuous optimization of the electronic map.

10. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 8.

Images