AI-based railway digital equipment model detection method

Abstract

The invention provides an AI-based railway digital equipment model detection method. The method comprises the following steps: constructing a feature coding system containing a railway digital equipment model; wherein the feature coding system comprises a model version, geometric features, parameter attributes and an association relationship of the railway digital equipment model; based on the feature coding system, integrating and aligning model data of each engineering stage of the railway digital equipment model to obtain unified model data; wherein the engineering stage model data comprises design stage model data, construction stage model data and operation and maintenance stage model data; based on the unified model data, performing multi-dimensional detection on the railway digital equipment model by using an AI technology to obtain a detection result so as to ensure the integrity and accuracy of the model; based on a pre-created knowledge base, recommending an optimization scheme for the detection result, and perfecting the knowledge base through continuous learning; and the knowledge base comprises a plurality of detection rules, so that problems in the model can be quickly found and optimization suggestions can be provided.

Description

Technical Field

[0001] This invention relates to the field of railway engineering inspection technology, and in particular to an AI-based method for inspecting railway digital equipment models. Background Technology

[0002] With the widespread application of Building Information Modeling (BIM) technology in railway engineering, digital modeling of railway equipment has achieved full lifecycle coverage. However, there are often discrepancies between the intended use of BIM models and the actual equipment.

[0003] Currently, BIM model inspection mainly relies on engineers comparing design parameters such as component dimensions and material properties. This manual verification method is extremely inefficient, susceptible to subjective interference, and has a high error rate. Furthermore, during railway operation, existing inspection technologies cannot promptly detect model updates (such as equipment modifications and parameter adjustments), which can lead to the inability to update and iterate the digital equipment model in relevant operation and maintenance systems in a timely manner. Summary of the Invention

[0004] This invention provides an AI-based method for detecting railway digital equipment models, which solves the technical problem that existing technologies cannot quickly and accurately verify railway digital equipment models.

[0005] On one hand, this invention provides an AI-based method for detecting railway digital equipment models, including: Construct a feature coding system that includes a railway digital equipment model; wherein the feature coding system includes the model version, geometric features, parameter attributes, and association relationships of the railway digital equipment model; Based on the aforementioned feature coding system, the model data of each engineering stage of the railway digital equipment model are integrated and aligned to obtain unified model data; wherein, the engineering stage model data includes design phase model data, construction phase model data, and operation and maintenance phase model data; Based on the unified model data, AI technology is used to perform multi-dimensional detection on the railway digital equipment model to obtain detection results, thereby ensuring the integrity and accuracy of the model; wherein, the multi-dimensional detection includes detection of geometric shape accuracy, parameter attribute rationality, and correlation correctness; Based on a pre-created knowledge base, optimization schemes are recommended for the detection results, and the knowledge base is continuously improved through learning; wherein, the knowledge base includes railway digital model creation standards, detection standards, design specifications, and maintenance rules.

[0006] According to the present invention, an AI-based railway digital equipment model detection method is provided, wherein the model data of each engineering stage of the railway digital equipment model are integrated and aligned based on the feature coding system to obtain unified model data, including: Based on the model version in the feature coding system, redundant and erroneous data in the model data of each engineering stage are eliminated; A semantic alignment rule base is established using the parameter attributes in the feature coding system to realize the semantic mapping and matching of parameter attributes of model data at each engineering stage; Combining the geometric features and correlations of the feature coding system, the iterative nearest point algorithm is used to perform geometric feature matching of cross-stage models, and the matching accuracy is controlled by the error threshold defined by the coding. Based on the spatiotemporal association rules of the feature coding system, the model data of each engineering stage are unified to the standard spatiotemporal reference through timestamp calibration and spatial coordinate transformation.

[0007] According to the present invention, a railway digital equipment model detection method based on AI is provided, wherein the railway digital equipment model is subjected to multi-dimensional detection using AI technology based on the unified model data to obtain detection results, including: The integrity of the unified model data is detected using a graph neural network; Geometric features of the unified model data are extracted using a convolutional neural network and compared with a standard feature library. An attention mechanism is used to identify discrepancies in the dynamic updating of unified model data. Compare the design parameters of the unified model data with the actual parameters during the operation and maintenance period, and mark abnormal data.

[0008] According to the present invention, a railway digital equipment model detection method based on AI, wherein the method recommends optimization schemes based on a pre-created knowledge base and continuously improves the knowledge base through learning, includes: Based on the type of deviation in the test results, match predefined maintenance strategies from the knowledge base; Dynamically adjust maintenance strategies using reinforcement learning algorithms; The maintenance strategies adopted by users and their actual effects are fed back to the knowledge base, and multi-dimensional detection and maintenance strategies are updated through supervised learning.

[0009] According to the present invention, an AI-based method for detecting railway digital equipment models includes constructing a feature encoding system containing railway digital equipment models, comprising: Based on the ISO 19650 standard, an ant colony algorithm was used to construct a feature coding system that includes a railway digital equipment model.

[0010] According to the AI-based railway digital equipment model detection method provided by the present invention, the construction of a feature coding system including the railway digital equipment model further includes: Based on the actual application scenarios and business needs of railway digital equipment, extension rules for the feature coding system are defined; wherein, the extension rules include dynamically adding new feature dimensions and coding rules without changing the existing coding system structure; Establish a version management mechanism for the feature coding system to record the content and reasons for each change to the coding system.

[0011] According to the present invention, a railway digital equipment model detection method based on AI, based on the model version in the feature coding system, removes redundant and erroneous data from the model data at each engineering stage, including: Version identification matching is performed on the model data at each stage of the project to identify duplicate and inconsistent data between different versions; Based on the priority rules of model versions, retain higher version data or verified correct data, and delete lower version or incorrect data; A data consistency verification algorithm is used to detect and remove abnormal data that does not conform to the characteristics of the model version.

[0012] According to the present invention, an AI-based railway digital equipment model detection method is provided, which utilizes parameter attributes in a feature coding system to establish a semantic alignment rule base, thereby realizing the semantic mapping and matching of parameter attributes of model data at each engineering stage, including: Extract semantic information of parameter attributes from model data at each engineering stage; wherein, the semantic information includes the parameter's name, type, unit, and range; A semantic alignment rule base is established based on the semantic information of parameter attributes; wherein, the semantic alignment rule base includes synonyms, near-synonyms and unit conversion rules between parameters; Natural language processing is used to perform semantic parsing on the parameter attributes of model data at each stage of the project, and the parsing results are obtained. The alignment rule base is based on the parsing results to achieve mapping and matching of parameter attributes.

[0013] According to the present invention, a railway digital equipment model detection method based on AI combines the geometric features and correlation relationships of a feature coding system, employs an iterative nearest-point algorithm for geometric feature matching of cross-stage models, and controls the matching accuracy through an error threshold defined by the coding, including: Based on the correlation in the feature coding system, determine the corresponding points for geometric feature matching between cross-stage models; The iterative nearest point algorithm is applied to match the geometric features of the cross-stage model, and the point cloud distance error and geometric shape deviation in the matching process are controlled by an error threshold defined by encoding.

[0014] On the other hand, the present invention also provides an AI-based railway digital equipment model detection system, comprising: The coding module is used to construct a feature coding system containing railway digital equipment models; wherein, the feature coding system includes the model version, geometric features, parameter attributes, and association relationships of the railway digital equipment models; The alignment module is used to integrate and align the model data of each engineering stage of the railway digital equipment model based on the feature coding system to obtain unified model data; wherein, the engineering stage model data includes design phase model data, construction phase model data and operation and maintenance phase model data; The detection module is used to perform multi-dimensional detection on the railway digital equipment model based on the unified model data and using AI technology to obtain detection results, so as to ensure the integrity and accuracy of the model; wherein, the multi-dimensional detection includes the detection of geometric shape accuracy, parameter attribute rationality, and correlation correctness; An optimization module is used to recommend optimization schemes for the detection results based on a pre-created knowledge base, and to continuously improve the knowledge base through learning; wherein, the knowledge base includes railway digital model creation standards, detection standards, design specifications, and maintenance rules.

[0015] This invention provides an AI-based method for detecting railway digital equipment models. By constructing a feature encoding system that includes railway digital equipment models, integrating and aligning model data from various engineering stages, utilizing AI technology for multi-dimensional detection, and recommending optimization solutions based on a knowledge base, this method can efficiently ensure the integrity and accuracy of railway digital equipment models, quickly identify problems in the models, and provide optimization suggestions. This improves the digital management level of railway engineering, reduces operation and maintenance costs, and enhances the operational reliability of railway equipment. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the AI-based railway digital equipment model detection method provided in an embodiment of the present invention. Figure 2This is a schematic diagram of the feature coding system provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the association relationships of the railway digital equipment model provided in the embodiments of the present invention; Figure 4 This is a schematic diagram of the integrity detection of the railway digital equipment model provided in an embodiment of the present invention; Figure 5 This is a schematic diagram visualizing the detection results of the railway digital equipment model provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of the AI-based railway digital equipment model detection system provided in an embodiment of the present invention; Figure 7 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0019] Figure 1 This is a flowchart illustrating the AI-based railway digital equipment model detection method provided in this embodiment of the invention. AI stands for Artificial Intelligence.

[0020] See Figure 1 The AI-based railway digital equipment model detection method may include the following steps.

[0021] Step 101: Construct a feature coding system that includes the railway digital equipment model; wherein, the feature coding system includes the model version, geometric features, parameter attributes, and correlation relationships of the railway digital equipment model.

[0022] In this step, "model version" refers to the version information of the railway digital equipment model at different design, construction, and operation and maintenance stages. For example, the model version initially created during the design phase can be defined as "V1.0," the version modified during the construction phase based on actual site conditions is "V1.1," and the version updated during the operation and maintenance phase after equipment modifications is "V2.0." Geometric features refer to the shape, size, and location of the railway digital equipment model. Geometric features describe the spatial attributes of the equipment, ensuring the geometric accuracy and consistency of the model; for example, the length, width, height, and shape of the equipment. Parameter attributes refer to various parameter information of the railway digital equipment model, such as material properties, performance parameters, and functional parameters. Parameter attributes describe the functional and performance characteristics of the equipment, ensuring that the model's parameters are consistent with the actual equipment; for example, the material, strength, and durability of the equipment. Association relationships refer to the connection and dependency relationships between different components in the railway digital equipment model. Association relationships describe the logical relationships between various components within the equipment, ensuring the integrity and consistency of the model; for example, the assembly relationships and connection sequence of the equipment. For example, the encoding format can be “ABC-001-G0123-P0045-R007”, where “ABC-001” represents the model version, “G0123” represents the geometric features, “P0045” represents the parameter attributes, and “R007” represents the association relationships. Generally, the model version can be 8 bits, geometric features can be 16 bits, parameter attributes can be 24 bits, and association relationships can be 16 bits.

[0023] The feature coding system can be the EBS coding system; for details, please refer to [link / reference needed]. Figure 2 The EBS coding system includes equipment type coding, specification coding, location coding, and function coding. Equipment type coding can use alphanumeric combinations to identify the equipment type, such as TXJ representing a signal controller. Specification coding can use a combination of numbers or characters to describe the specifications, such as 001 representing a specific model. Location coding can use geographic coordinates or regional codes to represent the location, such as E01 representing a specific area. Function coding can use alphanumeric combinations to specify the function, such as S01 representing signal control. The EBS coding system has the capability to achieve model-level code matching within seconds using an ant colony algorithm, with a code matching error rate strictly controlled to within 0.1%.

[0024] Step 102: Based on the feature coding system, integrate and align the model data of each engineering stage of the railway digital equipment model to obtain unified model data; among which, the engineering stage model data includes the design phase model data, the construction phase model data, and the operation and maintenance phase model data.

[0025] Step 102 may include: Step 1: Based on the model version in the feature coding system, eliminate redundant and erroneous data in the model data of each engineering stage.

[0026] In this step, version identification matching is performed on the model data of each engineering stage to identify duplicate data between different versions (e.g., a device is modeled repeatedly in multiple versions) and inconsistent data (e.g., a parameter is assigned different values ​​in different versions). Based on model version priority rules (e.g., higher version priority or verified version priority), retain higher version data or verified correct data, and delete lower version or incorrect data. Data consistency verification algorithms can be used to detect and remove erroneous data that does not conform to the characteristics of the model version.

[0027] Step 2: Utilize the parameter attributes in the feature coding system to establish a semantic alignment rule base, thereby realizing the semantic mapping and matching of parameter attributes of model data at each engineering stage.

[0028] In this step, semantic information of parameter attributes in the model data of each engineering stage is extracted, including the parameter name, type, unit, range, etc. For example, the "length" parameter of a certain device may be defined as "length" during the design phase, while it may be defined as "length_value" during the construction phase.

[0029] Based on the semantic information of parameter attributes, a semantic alignment rule base is established. This rule base includes synonyms, near-synonyms, and unit conversion rules between parameters. For example, "length" and "length_value" may be identified as synonyms, while there is a unit conversion relationship between "meter" and "millimeter".

[0030] Natural Language Processing (NLP) technology is used to semantically analyze the parameter attributes of model data at each engineering stage using methods such as lexical analysis, syntactic analysis, and semantic role labeling. For professional terms in the railway engineering field, a domain-specific thesaurus is constructed to optimize the training data of the NLP model and improve the accuracy of the analysis.

[0031] Based on the parsing results and the semantic alignment rule base, the mapping and matching of parameter attributes are realized. For example, the "length" parameter in the design phase is matched with the "length_value" parameter in the construction phase to ensure that they are semantically consistent.

[0032] Step 3: Combining the geometric features and correlations of the feature coding system, the iterative nearest point algorithm is used to perform geometric feature matching of the cross-stage model, and the matching accuracy is controlled by the error threshold defined by the coding.

[0033] In this step, based on the correlation in the feature coding system, the geometric feature matching points between cross-stage models are determined. For example, the geometric feature matching points between a certain piece of equipment in the design phase model and the corresponding piece of equipment in the construction phase model are determined through the correlation.

[0034] An iterative nearest-point algorithm is applied to match the geometric features of the cross-stage model, and a defined error threshold is used to control the point cloud distance error and geometric shape deviation during the matching process. For example, an error threshold of ±2mm is set to ensure that the matching accuracy meets the requirements.

[0035] Figure 3 This includes identification information (ID), geometric information (G), technical information (T), project information (P), and production information (MF). This information is organized in tabular form, with each attribute having a corresponding name and value. For example, identification information (ID) includes basic information such as the equipment's name, type, and IFD code. Geometric information (G) includes descriptions of the equipment's geometric features, such as length, width, and height. Technical information (T) includes technical parameters such as rated capacity, primary side rated voltage, secondary side rated voltage, primary side rated current, secondary side rated current, and rated frequency. Project information (P) involves project name, project stage, railway grade, owner information, designer information, and contractor information. Production information (MF) includes price, manufacturer, supplier, service life, maintenance cycle, and contact information.

[0036] Step 4: Based on the spatiotemporal association rules of the feature coding system, the model data of each engineering stage are unified to the standard spatiotemporal reference through timestamp calibration and spatial coordinate transformation.

[0037] In this step, the spatiotemporal association rules of the feature coding system are used to describe the temporal and spatial relationships of railway digital equipment models. For example, the timestamps and spatial coordinates of a certain piece of equipment may differ at different stages, requiring unification through spatiotemporal association rules. The timestamps of the model data at each engineering stage are calibrated to ensure temporal consistency. For example, time synchronization techniques and interpolation methods are used to unify data from different data sources to the same time scale. The spatial coordinates of the model data at each engineering stage are transformed, mapping the data to a unified spatial coordinate system. For example, coordinate transformation algorithms are used to convert the spatial coordinates of the design phase model to a coordinate system consistent with the construction phase model. Finally, the model data at each engineering stage are unified to a standard spatiotemporal reference, providing a complete and accurate data foundation for subsequent model testing and analysis.

[0038] The above steps significantly improve the efficiency and accuracy of multi-source model data fusion by eliminating redundant and erroneous data through model version management, achieving parameter attribute matching using a semantic alignment rule library, performing geometric feature matching with the iterative nearest point algorithm, and unifying data benchmarks through spatiotemporal association rules. This provides a high-quality and consistent data foundation for subsequent detection work, effectively avoids detection errors caused by data inconsistency, and enhances the level of refinement in railway digital equipment model management.

[0039] A multi-format BIM parser can be used, supporting common BIM formats such as Revit, Bentley, and IFC, and has corresponding parsing algorithms and interfaces, enabling it to parse BIM files of different formats into a unified data format.

[0040] Step 103: Based on unified model data, use AI technology to perform multi-dimensional inspections on the railway digital equipment model to obtain inspection results, ensuring the integrity and accuracy of the model. Multi-dimensional inspections may include checks on geometric accuracy, parameter attribute rationality, and the correctness of correlation relationships.

[0041] Step 103 may include: Step 1: Use graph neural networks to check the integrity of the unified model data (analyze the correctness of the topology and relationships).

[0042] The Graph Neural Network (GNN) used in this step is a deep learning model specifically designed for processing graph-structured data. It can effectively analyze the topology and relationships in railway digital equipment models. By using the GNN to check the integrity of the model data, including analyzing whether the connections between devices conform to design specifications and whether there are any broken or redundant connections, the model's topology and relationships are ensured to be compliant, preventing equipment malfunctions caused by connection errors.

[0043] Figure 4 The interface for integrity detection algorithm is displayed, divided into a left-hand list, a middle section, and a right-hand section. The left-hand list shows various components of the device, such as the cabinet, automatic control unit, power board, and fault indicator light.

[0044] The middle section displays detailed information about the selected component, including classification and attribute detection, geometric and spatial detection, etc. The right side displays the dimensions (length, width, and height) of the space occupied by the model. This interface provides an intuitive way to check and verify the integrity and spatial layout of the various components of the railway digital equipment model, ensuring the model's accuracy and consistency.

[0045] Step 2: Use a convolutional neural network to extract the geometric features of the unified model data and compare them with a standard feature library (to verify the geometric parameter error, that is, to detect the accuracy of the geometric shape).

[0046] The convolutional neural network (CNN) in this step can automatically extract key features from images or geometric data, such as shape, texture, and edges. Through multiple convolutional and pooling layers, the CNN can capture the geometric features of railway digital equipment models and compare them with a standard feature library to verify whether the geometric parameters of the model meet the design requirements, ensuring that the size and shape accuracy of the equipment are within the allowable error range.

[0047] Step 3: Use an attention mechanism to identify differences in the dynamic updates of the unified model data (implement version tree difference detection).

[0048] The attention mechanism in this step enables the model to focus on key information when processing large amounts of data, improving detection efficiency. By using the model version tree, it identifies differences between different versions, ensuring consistency of the model during dynamic updates. It can promptly detect changes in model updates and avoid operational problems caused by version differences.

[0049] Step 4: Compare the design parameters of the unified model data with the actual parameters during the operation and maintenance period, and mark abnormal data (detection of the rationality of parameter attributes).

[0050] In this step, the parameters in the design phase are compared with the actual parameters in the operation and maintenance phase to check for any deviations. Abnormal data that does not meet the design requirements or the actual operating status (deviations exceeding the preset deviation threshold) are marked to ensure the accuracy and consistency of the model parameters.

[0051] Step 104: Based on a pre-created knowledge base, recommend optimization schemes for the test results, and continuously improve the knowledge base through learning. The knowledge base may include railway digital model creation standards, testing standards, design specifications, maintenance rules, etc.

[0052] Specifically, the railway digital model creation standard provides specifications for model creation, ensuring initial model quality. Inspection results are referenced to the creation standard to determine whether the model meets initial construction requirements, thus providing a basis for optimization. The inspection standard clarifies the specific criteria and indicators for model inspection, determining the accuracy of the inspection results and serving as a key basis for evaluating model quality. The design specification stipulates the technical parameters and requirements for the model design phase. Inspection results are compared with the design specification; if deviations in design parameters are found, optimization solutions are recommended accordingly. Maintenance rules guide the maintenance and updates of the model during the operation and maintenance phase. Inspection results are used to evaluate the maintenance status of the model during the operation and maintenance phase based on the maintenance rules, identifying and resolving problems.

[0053] Step 104 may include: Step 1: Based on the type of deviation in the test results, match predefined maintenance strategies from the knowledge base; In this step, the knowledge base pre-stores various maintenance strategies for different types of deviations. The system will retrieve the corresponding predefined maintenance strategy from the knowledge base based on the deviation type identified in the detection results (such as geometric deviation, parameter deviation, topological deviation, etc.).

[0054] Step 2: Dynamically adjust the maintenance strategy using reinforcement learning algorithms; In this step, the reinforcement learning algorithm dynamically adjusts the maintenance strategy to adapt to different detection results by simulating the decision-making process in the environment. The system optimizes the parameters of the maintenance strategy based on actual detection conditions and historical data, making it more suitable for the current model state.

[0055] Step 3: Feed the user-adopted maintenance strategies and their actual effects into the knowledge base, and update the multi-dimensional detection and maintenance strategies through supervised learning.

[0056] In this step, the maintenance strategies adopted by users in actual applications and their implementation effects are recorded. Through supervised learning algorithms, the recorded feedback information is used to update the detection rules and maintenance strategies in the knowledge base, thereby continuously optimizing the knowledge base.

[0057] The test results can also be visualized, such as... Figure 5 As shown, the left-hand list displays each item inspected and its corresponding results. The middle section displays detailed inspection information, including inspection standards and results. The right-hand section presents a visual view of a 3D model, allowing users to rotate, zoom, and view different parts of the model. Inspection results can be visually displayed on the BIM model by establishing relationships, and different colors, icons, or labels can be used to distinguish fault types and states.

[0058] In this embodiment, by constructing a feature coding system that includes railway digital equipment models, integrating and aligning model data from each engineering stage, utilizing AI technology for multi-dimensional detection, and recommending optimization schemes based on a knowledge base, the integrity and accuracy of railway digital equipment models can be efficiently ensured, problems in the models can be quickly identified, and optimization suggestions can be provided. This improves the level of digital management of railway engineering, reduces operation and maintenance costs, and enhances the operational reliability of railway equipment.

[0059] In one embodiment of this specification, a feature coding system incorporating a railway digital equipment model is constructed, including: Based on the ISO 19650 standard, an ant colony algorithm was used to construct a feature coding system that includes a railway digital equipment model.

[0060] In this embodiment, the ISO 19650 standard is adopted to ensure the standardization and compatibility of the feature coding system, enabling it to interface with existing BIM management processes. When constructing the feature coding system, the ant colony algorithm is used to efficiently generate and optimize codes. By simulating the search behavior of ants in the coding space, it quickly finds the optimal coding path, thereby improving coding efficiency and accuracy.

[0061] In one embodiment of this specification, constructing a feature coding system that includes a railway digital equipment model further includes: Based on the actual application scenarios and business needs of railway digital equipment, extension rules for the feature coding system are defined; among them, the extension rules include dynamically adding new feature dimensions and coding rules without changing the existing coding system structure. Establish a version management mechanism for the feature coding system to record the content and reasons for each change to the coding system.

[0062] In this embodiment, the application scenarios and business requirements of railway digital equipment models are diverse, and different feature dimensions may need to be considered in different scenarios. For example, in railway bridge inspection, additional attention may be needed to structural strength and durability parameters; while in railway signaling systems, signal transmission efficiency and fault response time may need to be considered. To meet these diverse scenario requirements, the feature coding system needs to be scalable, capable of dynamically adding new feature dimensions and coding rules without changing the existing coding system architecture. A version management mechanism is used to record the changes and reasons for changes to the coding system. This helps to trace historical versions and understand the background and purpose of changes after the coding system is updated. The version management mechanism may include functions such as version number allocation, change log recording, and comparison of differences between versions. For example, each time the coding system is updated, a new version number is generated, and the specific content and reason for the update are recorded.

[0063] Based on the same general inventive concept, this invention also protects an AI-based railway digital equipment model detection system, such as... Figure 6 As shown, Figure 6 This is a schematic diagram of the structure of the AI-based railway digital equipment model detection system provided in an embodiment of the present invention. The AI-based railway digital equipment model detection system provided by the present invention will be described below. The AI-based railway digital equipment model detection system described below can be referred to in correspondence with the AI-based railway digital equipment model detection method described above.

[0064] The AI-based railway digital equipment model detection system includes an encoding module 601, an alignment module 602, a detection module 603, and an optimization module 604.

[0065] The encoding module 601 is used to construct a feature encoding system containing a railway digital equipment model; wherein, the feature encoding system includes the model version, geometric features, parameter attributes, and association relationships of the railway digital equipment model; The alignment module 602 is used to integrate and align the model data of each engineering stage of the railway digital equipment model based on the feature coding system to obtain unified model data; wherein, the engineering stage model data includes design phase model data, construction phase model data and operation and maintenance phase model data; The detection module 603 is used to perform multi-dimensional detection on the railway digital equipment model based on the unified model data and using AI technology to obtain detection results, so as to ensure the integrity and accuracy of the model; wherein, the multi-dimensional detection includes the detection of geometric shape accuracy, parameter attribute rationality, and correlation correctness; The optimization module 604 is used to recommend optimization schemes for the detection results based on a pre-created knowledge base, and to continuously improve the knowledge base through learning; wherein, the knowledge base includes railway digital model creation standards, detection standards, design specifications, and maintenance rules.

[0066] Figure 7 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention.

[0067] like Figure 7 As shown, the electronic device may include a processor 710, a communications interface 720, a memory 730, and a communication bus 740. The processor 710, communications interface 720, and memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions from the memory 730 to execute an AI-based railway digital equipment model detection method.

[0068] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0069] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the AI-based railway digital equipment model detection method provided by the above methods.

[0070] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the AI-based railway digital equipment model detection method provided by the above methods.

[0071] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0072] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the corresponding technical solutions to deviate from the essence and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An AI-based method for detecting railway digital equipment models, characterized in that, include: Construct a feature coding system that includes a railway digital equipment model; wherein the feature coding system includes the model version, geometric features, parameter attributes, and association relationships of the railway digital equipment model; Based on the aforementioned feature coding system, the model data of each engineering stage of the railway digital equipment model are integrated and aligned to obtain unified model data; wherein, the engineering stage model data includes design phase model data, construction phase model data, and operation and maintenance phase model data; Based on the unified model data, AI technology is used to perform multi-dimensional detection on the railway digital equipment model to obtain detection results, thereby ensuring the integrity and accuracy of the model; wherein, the multi-dimensional detection includes detection of geometric shape accuracy, parameter attribute rationality, and correlation correctness; Based on a pre-created knowledge base, optimization schemes are recommended for the detection results, and the knowledge base is continuously improved through learning; wherein, the knowledge base includes railway digital model creation standards, detection standards, design specifications, and maintenance rules.

2. The AI-based railway digital equipment model detection method according to claim 1, characterized in that, Based on the feature coding system, the model data of each engineering stage of the railway digital equipment model are integrated and aligned to obtain unified model data, including: Based on the model version in the feature coding system, redundant and erroneous data in the model data of each engineering stage are eliminated; A semantic alignment rule base is established using the parameter attributes in the feature coding system to realize the semantic mapping and matching of parameter attributes of model data at each engineering stage; Combining the geometric features and correlations of the feature coding system, the iterative nearest point algorithm is used to perform geometric feature matching of cross-stage models, and the matching accuracy is controlled by the error threshold defined by the coding. Based on the spatiotemporal association rules of the feature coding system, the model data of each engineering stage are unified to the standard spatiotemporal reference through timestamp calibration and spatial coordinate transformation.

3. The AI-based railway digital equipment model detection method according to claim 1, characterized in that, Based on the unified model data, AI technology is used to perform multi-dimensional detection on the railway digital equipment model to obtain detection results, including: The integrity of the unified model data is detected using a graph neural network; Geometric features of the unified model data are extracted using a convolutional neural network and compared with a standard feature library. An attention mechanism is used to identify discrepancies in the dynamic updating of unified model data. Compare the design parameters of the unified model data with the actual parameters during the operation and maintenance period, and mark abnormal data.

4. The AI-based railway digital equipment model detection method according to claim 1, characterized in that, The process of recommending optimization schemes for the detection results based on a pre-created knowledge base, and continuously improving the knowledge base through learning, includes: Based on the type of deviation in the test results, match predefined maintenance strategies from the knowledge base; Dynamically adjust maintenance strategies using reinforcement learning algorithms; The maintenance strategies adopted by users and their actual effects are fed back to the knowledge base, and multi-dimensional detection and maintenance strategies are updated through supervised learning.

5. The AI-based railway digital equipment model detection method according to claim 1, characterized in that, The construction of the feature coding system containing railway digital equipment models includes: Based on the ISO 19650 standard, an ant colony algorithm was used to construct a feature coding system that includes a railway digital equipment model.

6. The AI-based railway digital equipment model detection method according to claim 1, characterized in that, The construction of the feature coding system including the railway digital equipment model also includes: Based on the actual application scenarios and business needs of railway digital equipment, extension rules for the feature coding system are defined; wherein, the extension rules include dynamically adding new feature dimensions and coding rules without changing the existing coding system structure; Establish a version management mechanism for the feature coding system to record the content and reasons for each change to the coding system.

7. The AI-based railway digital equipment model detection method according to claim 2, characterized in that, Based on the model version in the feature coding system, redundant and erroneous data in the model data of each engineering stage are removed, including: Version identification matching is performed on the model data at each stage of the project to identify duplicate and inconsistent data between different versions; Based on the priority rules of model versions, retain higher version data or verified correct data, and delete lower version or incorrect data; A data consistency verification algorithm is used to detect and remove abnormal data that does not conform to the characteristics of the model version.

8. The AI-based railway digital equipment model detection method according to claim 2, characterized in that, A semantic alignment rule base is established using parameter attributes in the feature coding system to achieve semantic mapping and matching of parameter attributes of model data at each engineering stage, including: Extract semantic information of parameter attributes from model data at each engineering stage; wherein, the semantic information includes the parameter's name, type, unit, and range; A semantic alignment rule base is established based on the semantic information of parameter attributes; wherein, the semantic alignment rule base includes synonyms, near-synonyms and unit conversion rules between parameters; Natural language processing is used to perform semantic parsing on the parameter attributes of model data at each stage of the project, and the parsing results are obtained. The alignment rule base is based on the parsing results to achieve mapping and matching of parameter attributes.

9. The AI-based railway digital equipment model detection method according to claim 2, characterized in that, Combining the geometric features and correlations of the feature coding system, an iterative nearest-point algorithm is used for geometric feature matching across stages of the model. Matching accuracy is controlled by an error threshold defined in the coding, including: Based on the correlation in the feature coding system, determine the corresponding points for geometric feature matching between cross-stage models; The iterative nearest point algorithm is applied to match the geometric features of the cross-stage model, and the point cloud distance error and geometric shape deviation in the matching process are controlled by an error threshold defined by encoding.

10. An AI-based railway digital equipment model inspection system, characterized in that, include: The coding module is used to construct a feature coding system containing railway digital equipment models; wherein, the feature coding system includes the model version, geometric features, parameter attributes, and association relationships of the railway digital equipment models; The alignment module is used to integrate and align the model data of each engineering stage of the railway digital equipment model based on the feature coding system to obtain unified model data; wherein, the engineering stage model data includes design phase model data, construction phase model data and operation and maintenance phase model data; The detection module is used to perform multi-dimensional detection on the railway digital equipment model based on the unified model data and using AI technology to obtain detection results, so as to ensure the integrity and accuracy of the model; wherein, the multi-dimensional detection includes the detection of geometric shape accuracy, parameter attribute rationality, and correlation correctness; An optimization module is used to recommend optimization schemes for the detection results based on a pre-created knowledge base, and to continuously improve the knowledge base through learning; wherein, the knowledge base includes railway digital model creation standards, detection standards, design specifications, and maintenance rules.

Images