Urban rail transit vehicle configuration digital modeling system and method

Abstract

The application discloses a kind of urban rail transit system vehicle configuration digital modeling system and method, it is related to rail transit engineering technical field, including the following steps: S100, establish the global level baseline of vehicle configuration, generate unique topological fingerprint to each node in configuration level, and build the corresponding relationship of topological fingerprint and time anchor point in combination with node creation time, form the level signature library for abnormal structure comparison.The application realizes accurate identification and isolation to configuration loop by topological fingerprint identification, level signature verification, differential comparison and sandbox verification, and dynamically optimizes structure import order in combination with risk score and self-healing control, significantly improves the accuracy, controllability and system robustness of rail transit vehicle configuration data, and provides reliable support for cross-car type operation and maintenance.

Description

Technical Field

[0001] This invention relates to the field of rail transit engineering technology, specifically to a digital modeling system and method for vehicle configuration in urban rail transit systems. Background Technology

[0002] Digital modeling of vehicle configuration in urban rail transit systems refers to the digital representation and structured modeling of the hierarchical relationships of rail vehicles, from model type and carriages to equipment systems, equipment types, and components, under the requirements of full life-cycle vehicle management. It accurately maps the actual configuration of each vehicle into a digital model by constructing a unified equipment classification system, equipment list, equipment BOM, model structure, and vehicle equipment ledger. This modeling not only includes geometric and structural configurations but also incorporates multi-dimensional information such as equipment model, specifications, life cycle, and serial number, thus forming a complete digital profile of the vehicle and supporting standardized management and information exchange across vehicle models and lines.

[0003] The core value of this digital modeling lies in its provision of refined digital support for the inspection, operation, maintenance, and upgrading of rail transit vehicles. By combining vehicle configuration information with real-time operational data and health monitoring results collected by the IoT platform, it enables end-to-end tracking and dynamic management from individual components to the entire vehicle. For example, when a certain type of equipment malfunctions, the system can quickly locate the vehicle model, carriage, and installation location based on the modeling results, and retrieve the corresponding BOM and ledger information to achieve precise maintenance and spare parts management. Simultaneously, this digital modeling approach also lays the foundation for future digital twins and intelligent operation and maintenance, enabling urban rail transit systems to maintain efficient, transparent, and controllable operation even as they face expansion and increased complexity.

[0004] Existing technologies suffer from the following shortcomings: In current technologies, digital modeling of urban rail transit vehicle configurations typically relies on recursive associations of multi-level components to organize and manage the compositional relationships of equipment. However, due to the lack of effective constraints on hierarchical reference relationships, erroneous loop calls between components are easily generated during the modeling process. This type of problem is often difficult to detect during the initial data entry phase; the data appears complete on the surface. However, when the operation and maintenance system performs BOM expansion, hierarchical tracing, or automatic calls, it triggers infinite recursion, causing the system to enter a dead loop and directly leading to the collapse of the Vehicle Maintenance Information Management Platform (VMS). This hidden logical error not only seriously affects the integrity and reliability of the vehicle's electronic history but may also render maintenance decisions, lifespan predictions, and equipment ledger management ineffective, thus posing a significant threat to the operational safety and maintenance efficiency of rail transit.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a digital modeling system and method for vehicle configuration in urban rail transit systems to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a digital modeling method for vehicle configuration in urban rail transit systems, comprising the following steps:

[0008] S100: Establish a global hierarchical baseline for vehicle configuration, generate a unique topological fingerprint for each node in the configuration hierarchy, and construct the correspondence between the topological fingerprint and the time anchor point by combining the node creation time, forming a hierarchical signature library for anomaly structure comparison.

[0009] S200, based on the hierarchical signature library, performs structural difference branching kernels and uses the quantity conservation rule, hierarchical monotonic rule and parent-child information entropy threshold rule for parallel comparison to generate a list of nodes suspected of having topological loops;

[0010] S300, for the list of nodes suspected of having topological loops, adopts the reachability cone domain traversal method to expand the vehicle configuration level from the root node, detects the topological fingerprint duplication during the traversal process and marks it immediately, and constructs a complete evidence chain of topological loop links.

[0011] S400, based on the loop link topology, performs topology shrinkage on the loop substructure in an isolated sandbox environment, replaces the loop node with the structurally equivalent placeholder node, simulates the unfolding process and records the first position that cannot be terminated and the stack unfolding depth.

[0012] S500 inputs the location where termination is not possible and the stack expansion depth as halting indicators into the verification engine, activates the cross-level structural consistency detection mechanism, and performs risk scoring on the reference structure between the bill of materials level, vehicle structure and vehicle equipment configuration.

[0013] S600 activates a self-healing finite state control process based on risk scoring results, performs a freeze operation on structural nodes that reach the risk critical threshold, and adjusts the structural introduction sequence through a fractional-order perturbation propulsion mechanism to achieve structural stability optimization of the vehicle configuration.

[0014] Preferably, step S100 includes:

[0015] Structural data was collected for each type of urban rail transit vehicle, each car, each carriage, each functional device and its subordinate components, and a parent-child cascaded structural path was constructed from the whole vehicle level to the component level.

[0016] A topological fingerprint is generated for each node in the structural path. The topological fingerprint is generated by sequentially concatenating the node's full path level block, structural attribute block, and structural location block and then inputting the hash code to ensure its uniqueness in the data dimension, time dimension, and location dimension.

[0017] The topological fingerprint is bound to the creation time of the node to build a one-to-one mapping relationship between the topological fingerprint and the time anchor point, forming a structural time identifier;

[0018] The topological fingerprints and time anchors of all structural nodes are summarized and arranged according to vehicle number, carriage number, hierarchical path, fingerprint value and timestamp to construct a full hierarchical signature set covering the configuration structure of all urban rail transit vehicles.

[0019] Preferably, step S200 includes:

[0020] Extract the current structural snapshot of each urban rail transit train and compare it with the manufacturing structural template at the path level to identify newly added nodes, missing nodes, and path offset nodes.

[0021] Based on the comparison results, a quantity conservation check is performed on each parent node, and first-level suspected abnormal nodes with abnormal child node counts or path backtracking structures are marked.

[0022] Perform hierarchical monotonicity checks on first-level suspected abnormal nodes, identify path hierarchy decreases, parent-child path overlaps, or path reversal structures, and record the number of overlapping path segments.

[0023] Further perform parent-child information entropy difference analysis on suspected nodes, calculate the difference in attribute complexity between parent and child nodes and the information entropy inversion coefficient, and form a list of nodes suspected of having topological loops.

[0024] Preferably, step S300 includes:

[0025] Starting from the root node of the configuration tree of each urban rail transit vehicle, load the structure snapshot and initialize the path tracking list to ensure that the structure data is consistent with the hierarchical signature set.

[0026] The structure nodes are expanded layer by layer using a depth-first approach, and the topological fingerprint of the current node is compared with the historical fingerprints in the path tracing list one by one in each expansion operation;

[0027] If a fingerprint value that is exactly the same as the topological fingerprint of the current node is found in the path, the current path expansion is immediately interrupted and the duplicate path information is recorded.

[0028] A closed-loop path chain is constructed based on repeated fingerprint paths, recording the topological fingerprint value, structural path, device number, parent-child relationship and time anchor point of all nodes in the path, thus completing the generation of the closed-loop structural evidence chain.

[0029] Preferably, step S400 includes:

[0030] Extract the node sequence that forms a closed loop in the cyclic path chain of the topology, identify the fingerprint repeating nodes, and extract the closed loop structure segment.

[0031] In the simulation environment, the closed-loop structure segment is replaced with the equivalent structural placeholder node, and the logical mapping relationship between the placeholder structure and the original structure is established.

[0032] Simulate unfolding operations based on the execution path of the configuration root node, and record the first position where the structure unfolding cannot be terminated and the maximum unfolding depth of the stack during the structure unfolding process;

[0033] Abnormal behaviors during the simulation deployment are categorized and summarized to generate a structure deployment failure report that includes the closed-loop path number, fingerprint value, path depth, and the status of the placeholder node.

[0034] Preferably, step S500 includes:

[0035] A structural downtime index table is established based on the structural downtime index, and the integrity of the path text, topological fingerprint, belonging number and stack level in each record is checked.

[0036] Call the three structural information items: vehicle equipment configuration list, vehicle model structure standard template and vehicle bill of materials hierarchy table, extract the field content related to downtime indicators and establish a horizontal comparison relationship;

[0037] Structural consistency is compared based on four criteria: path consistency, device attribute differences, parent-child reference relationship stability, and hierarchical number offset, and structural difference judgment results are generated.

[0038] The structural difference results are converted into structural risk scores, which are then compiled into a structural risk score list. Structural nodes with scores higher than a set threshold are marked as high-risk nodes.

[0039] Preferably, step S600 includes:

[0040] Filter all structural nodes with risk scores not lower than the set threshold, generate a list of structural risk interventions, and construct a structural freeze status mapping table;

[0041] Before importing the structure, freeze all high-risk nodes in the structural risk intervention list, replace them with structural placeholder nodes, and block direct references in all structural paths.

[0042] Perform an import order adjustment operation based on a fractional-order perturbation propulsion mechanism on the remaining structural path queue, reorder the path loading order and insert structural buffer nodes to reduce path dependency coupling;

[0043] Based on the perturbation sorting results, structural loading simulation tests were conducted to verify the continuity and stability of the structural path. The execution plan was then imported and the unfreezing and recovery conditions for the frozen nodes were set.

[0044] A digital modeling system for vehicle configuration in urban rail transit systems includes a hierarchical signature generation module, a structural difference kernel module, a loop evidence chain extraction module, a sandbox topology verification module, a risk scoring engine module, and a structural self-healing control module.

[0045] The hierarchical signature generation module establishes a global hierarchical baseline for vehicle configuration, generates a unique topological fingerprint for each node in the configuration hierarchy, and constructs a correspondence between the topological fingerprint and the time anchor point by combining the node creation time, forming a hierarchical signature library for comparing abnormal structures.

[0046] The structural difference branching kernel module executes the structural difference branching kernel based on the hierarchical signature library. It uses the quantity conservation rule, the hierarchical monotonic rule, and the parent-child information entropy threshold rule for parallel comparison to generate a list of nodes suspected of having topological loops.

[0047] The loop evidence chain extraction module targets a list of nodes suspected of having topological loops. It uses a reachability cone domain traversal method to expand the vehicle configuration level from the root node. During the traversal, it detects and immediately marks the duplicate topological fingerprints, thus constructing a complete evidence chain of loop links in the topological structure.

[0048] The sandbox topology verification module, based on the loop link topology structure, performs topology shrinkage on the loop substructure in an isolated sandbox environment, replaces the loop nodes with structurally equivalent placeholder nodes, simulates the unfolding process, and records the first position where it cannot be terminated and the stack unfolding depth.

[0049] The risk scoring engine module inputs the location where termination is impossible and the stack expansion depth as shutdown indicators into the verification engine, activates the cross-level structural consistency detection mechanism, and performs risk scoring on the reference structure between the bill of materials level, vehicle model structure and vehicle equipment configuration.

[0050] The structural self-healing control module activates the self-healing finite state control process based on the risk score results, performs a freeze operation on structural nodes that reach the risk critical threshold, and adjusts the structural import sequence through a fractional-order perturbation propulsion mechanism to achieve structural stability optimization of the vehicle configuration.

[0051] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0052] This invention generates unique topological fingerprints and constructs a hierarchical signature library, enabling each structural node to possess a verifiable identity and enhancing structural integrity verification capabilities. Simultaneously, it introduces a structural difference core and information entropy comparison mechanism, allowing potential loop nodes to be located and isolated during the data entry stage. Furthermore, by combining topological evidence chains with topological shrinkage experiments in a sandbox environment, it can simulate the real configuration loading process and capture the location of structural anomalies. Finally, through risk scoring and self-healing state control, it dynamically adjusts the structural import order, effectively breaking path dependency loops and achieving proactive structural stability and continuous optimization. Overall, this solution significantly improves the accuracy, controllability, and system robustness of rail transit vehicle configuration data and the operating platform, providing a solid data support foundation for digital operation and maintenance and intelligent scheduling across lines and vehicle types. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0054] Figure 1 This is a flowchart of a digital modeling method for vehicle configuration in an urban rail transit system according to the present invention.

[0055] Figure 2 This is a schematic diagram of a module of a digital modeling system for vehicle configuration in an urban rail transit system according to the present invention. Detailed Implementation

[0056] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0057] This invention provides, for example Figure 1 The method for digital modeling of vehicle configuration in an urban rail transit system, as shown, includes the following steps:

[0058] S100: Establish a global hierarchical baseline for vehicle configuration, generate a unique topological fingerprint for each node in the configuration hierarchy, and construct the correspondence between the topological fingerprint and the time anchor point by combining the node creation time, forming a hierarchical signature library for anomaly structure comparison.

[0059] To address issues such as loop calls, recursive infinite loops, and platform crashes that may arise from existing configuration modeling techniques in practical operation and maintenance systems, the following specific configuration modeling implementation process is proposed. This process constructs a global hierarchical baseline for the vehicle configuration, assigns a unique topological fingerprint identifier to each node in the configuration structure, and binds it one-to-one with the configuration creation time, thereby forming a complete set of structural signatures. This provides a technical foundation for subsequent anomaly detection and structural consistency comparison. The implementation process consists of the following steps:

[0060] Based on data from actual urban rail transit vehicles in operation, structural data is collected for each vehicle type, each car, each carriage, each functional device, and its subordinate components to establish a multi-level hierarchical configuration description. This structure adopts a "parent-child cascade" organizational approach, starting from the overall vehicle level and progressively descending to the carriage level, equipment system level, equipment level, component level, and part level. Each level requires a clear mapping relationship between its superior and subordinate levels. For example, a metro train numbered T1001 contains six carriages, each containing several functional units such as the braking system, signaling system, onboard power supply, and air conditioning / ventilation system. Taking the air conditioning / ventilation system as an example, it may contain multiple sub-components such as condenser fans, heaters, air outlet components, and filter units. During the structure establishment process, a structural entry needs to be created for each actual device or component, recording its installation location, functional affiliation, number identifier, data acquisition interface number, and equipment type code. All structural entries must form a complete hierarchical path, for example: "T1001 train → Carriage 3 → Air conditioning and ventilation system → Condenser fan → Air outlet impeller assembly". This path must uniquely represent the exact structural location of the node within the entire vehicle, and duplicate names at the same level or ambiguous paths are not allowed. Based on this, a complete configuration map covering all vehicles, all carriages, and all equipment is constructed, forming a hierarchical structure dataset that spans the entire urban rail transit vehicle fleet.

[0061] For each node in the aforementioned structured dataset, a unique topological fingerprint is generated to identify its identity and position within the structural configuration. The topological fingerprint is generated as follows: First, the full path hierarchy of each node is expanded, for example, the aforementioned "train number / carriage number / system name / equipment name / component name," arranged sequentially using a fixed-length structure. Second, eight key attributes—functional attributes, equipment manufacturer name, equipment serial number, equipment installation interface number, communication protocol encoding, data acquisition port number, power supply method, and size specifications—are arranged in a fixed order to form a structural attribute block. Then, relational parameters such as the node's corresponding hierarchy depth value in the structural graph, parent node reference number, and sibling node sequence number are appended to the end, forming a structural location block. Finally, the path hierarchy block, structural attribute block, and structural location block are concatenated in a fixed order to generate a text sequence in a unified format. This text sequence is used as input and encoded using a deterministic hash algorithm (such as SHA256 but not for public-key encryption) to ensure that each structural unique fingerprint generation result is unique and collision-resistant across the data, time, and location dimensions. This generation method differs from the simplified approach in existing technologies that relies solely on device type or manufacturer number to establish configuration structures. By combining complete paths with multi-dimensional structural attributes, it creates more accurate and distinguishable topological identifiers, facilitating subsequent tracking and anomaly analysis.

[0062] The generated topological fingerprint is bound to the node's creation time, forming a one-to-one mapping between the topological fingerprint and the time anchor. The creation time refers to the time when the device or component is first registered in the system. This is typically achieved by the platform automatically recording the device's import timestamp, or by using the earliest of the device's manufacturing date, installation and commissioning date, and first data upload date as the final anchor. For devices with replacement records, a new anchor must be marked based on the replacement date provided in the device replacement document. To ensure the uniqueness and verifiability of the time data, time verification is performed by connecting to an asset ledger system or electronic record system to ensure the time anchor's source is authentic, there is no risk of rewriting, and it is consistent with the device's historical status. Once bound to the topological fingerprint, this time anchor constitutes the structural time identifier. For example, the condenser fan with serial number A01876 is installed in the air conditioning system of carriage 3 of train T1001 at 12:06:55 on March 18, 2023. The topological fingerprint of this device is FJX92K-WMG5-LK001-XR21, and the two form a binding relationship "FJX92K-WMG5-LK001-XR21@2023-03-18T12:06:55Z". All binding results must be organized according to the vehicle, carriage, and structural path hierarchy to form a list of structure and time comparisons. Each record in the list must be immutable and auditable for future system structure traceability, version rollback, and replacement detection.

[0063] After generating topological fingerprints and binding time anchors for all nodes, the fingerprint records of all nodes are summarized and organized in a unified format to construct a full-scale hierarchical signature set for the entire rail transit vehicle configuration data. This set adopts a multi-dimensional structure, with each signature indexed and arranged according to five items: "vehicle number / carriage number / hierarchical path / fingerprint value / timestamp." Duplicate fingerprint and time anchor combinations are not allowed in the set. For equipment that has been replaced multiple times, a multi-version snapshot method is used to retain historical structural information. During the establishment of this set, fingerprint uniqueness verification is performed simultaneously to check for anomalies such as fingerprint conflicts, path overlaps, and anchor overlaps globally, ensuring the integrity and logical consistency of the set. This set is ultimately used for multiple subsequent functional scenarios such as configuration data version management, structural change detection, and loop warning identification. Unlike existing technologies that only establish static equipment lists, the signature set constructed by this method has unified expression capabilities across equipment, carriages, and vehicle models, and can achieve structural-level linkage with vehicle electronic history systems and BOM list management systems, thus forming a unified data expression foundation with strong scalability and reliability.

[0064] The purpose of this step is to establish a fundamental data system for urban rail transit vehicle configurations that is unique, traceable, and structurally complete, solving the structural loop errors and data distortion problems caused by ambiguous node identification, chaotic hierarchical relationships, and missing time information in traditional modeling methods. By generating a unique topological fingerprint for each configuration node and establishing a time anchor point based on the node's creation time, a complete set of structural signatures is formed. This not only enables accurate identification at the node level but also ensures the clarity of structural reference relationships and the traceability of historical evolution. This mechanism fundamentally guarantees the consistency and controllability of configuration modeling at the data level, providing a reliable reference for subsequent anomaly detection, recursive path identification, equipment history comparison, and lifecycle management. This effectively supports the efficient operation and unified management of vehicle structural information in large-scale, multi-model, and multi-line environments.

[0065] S200, based on the hierarchical signature library, performs structural difference branching kernels and uses the quantity conservation rule, hierarchical monotonic rule and parent-child information entropy threshold rule for parallel comparison to generate a list of nodes suspected of having topological loops;

[0066] To effectively identify potential circular calls, hierarchical reference errors, or structural logic anomalies in the configuration structure, a structural differential kernel operation based on the hierarchical signature set needs to be performed after topological fingerprint generation and time anchor binding. This process meticulously compares the configuration data from multiple structural consistency dimensions, sequentially identifying a set of nodes with anomalies in terms of quantity structure, hierarchical direction, and information entropy gradient, and ultimately generating a list of structural nodes suspected of containing topological loops. The specific implementation method is as follows, consisting of the following steps:

[0067] Structural information for each urban rail transit train is extracted from the constructed hierarchical signature set and categorized based on vehicle number. A current structural snapshot of each train is then extracted. This snapshot, derived from the latest round of structural data collection, includes fields such as train number, carriage number, equipment system name, equipment type name, component name, part number, equipment serial number, equipment installation location number, parent reference path, topology fingerprint value, and time anchor point. Subsequently, the current structural snapshot is matched one-to-one with the initial manufacturing structural template of the corresponding train model, establishing a comparison relationship between each structural node's current state and baseline state. The structural template, serving as the comparison baseline, must originate from a complete structural list provided by the manufacturing end, covering all equipment to be installed and their standard hierarchical locations. This comparison operation employs a path-level comparison method, determining node correspondence by comparing the complete path string of the structural path text (e.g., "train number / T1001→carriage number / C3→power system→power conversion device→rectifier module"), and simultaneously determining equipment consistency through topology fingerprint value matching. The comparison results are recorded in groups by node, and newly added nodes, missing nodes, and path offset nodes are marked to provide data input for the next step of verification.

[0068] Based on the above comparison results, a quantity conservation check is performed on each parent node. The criterion for quantity conservation is: if a parent node is recorded in the structural template as having 3 child nodes, but only 2 or more child nodes exist in the current snapshot, then the node is judged to violate the quantity conservation principle. For example, the power supply system of carriage "C3" should include 3 components in the structural template: a voltage conversion unit, a current monitoring unit, and a fault disconnection device. However, if only the voltage conversion unit and the current monitoring unit are found in the current snapshot, then the fault disconnection device is missing. Conversely, if there is an additional device with a similar structural path to the existing node but a different number, then it is an excess structure. For both of the above cases, the actual child node topology fingerprint must be recorded and compared with the parent node path segment to identify whether there is a repetition or cyclic nesting of structural path segments. For example, if the child node path contains "...→power system→voltage conversion unit→power system", then there is a possibility of a path reversal structure. The parent and child nodes marked above will be temporarily classified as first-level suspected abnormal structures and will proceed to the next comparison process.

[0069] For all suspected first-level abnormal structural nodes, a hierarchical monotonicity check is performed. The hierarchical monotonicity check means that in the structural expansion path, the hierarchical numbers of all paths from the root node to the leaf node should be incremental or strictly progressive, and the structural path direction should always expand from top to bottom, without upward references or horizontal cross-references. During implementation, the position of each node in the path is numbered, and the difference between its parent node number and the parent node number is calculated. If the hierarchical number of a child node is less than or equal to that of its parent node, or if a parent path identifier appears in its structural path (i.e., the child path is a subset of the parent path), then a path backtracking phenomenon is considered to have occurred. For example, in train number "T1003", if the electrical equipment path is "...→control unit→power interface→control unit", then the control unit is referenced by itself, constituting a path reversal. In this case, the starting node, backtracking node, path length, and backtracking level are recorded, and the number of overlapping path segments is calculated. Any path with more than two consecutive structural levels of repeated segments is considered to constitute a structural level anomaly and is included in the set of secondary suspected anomaly nodes.

[0070] In the aforementioned set of nodes marked as potentially abnormal during the verification process, a parent-child information entropy difference analysis is further performed to identify implicitly abnormal structures that appear error-free in terms of surface path and quantity but exhibit unreasonable expansion in information complexity. The information entropy difference analysis method involves first extracting the number of structural attribute items contained in the parent node, such as the number of attribute fields, data interface items, and control relationships; then extracting the average number of attribute items for all its child nodes and summing them. If the attribute complexity of the child node set is significantly greater than that of the parent node, and there is no hierarchical decay relationship between the parent and child nodes, then the node is determined to have structural information inversion. For example, a monitoring unit may only contain basic status feedback, but its child nodes may contain as many as 12 data structures, 6 control protocols, and 5 operating strategy interfaces; in this case, information generalization has occurred in the structure. This is especially true when such structures simultaneously exhibit hierarchical number reversal, path backflow, or duplicate device sequences, making them more likely to become entry points for recursive calls. These structures are compiled according to their unique path identifiers to form a complete list of nodes suspected of containing topological loops. The list records each node's path number, parent node number, number of child nodes, path overlap, attribute complexity ratio, information entropy inversion coefficient, structural fingerprint value, and time anchor information. This list will serve as a crucial input for subsequent path reachability analysis and the generation of closed-loop evidence chains.

[0071] The purpose of this step is to systematically differentiate and validate the established hierarchical signature set of urban rail transit vehicle configurations, accurately identifying potential topological loops and hierarchical reference anomalies within the structure. By performing quantity conservation checks, missing or redundant child nodes in the equipment structure can be detected; by hierarchical monotonicity checks, structural directional issues such as path reversal and parent-child reference errors can be identified; further, through parent-child information entropy difference analysis, potential erroneous nodes with seemingly reasonable structures but anomalous expansions in attribute complexity can be identified. This step not only cross-validates the configuration data from three dimensions—structural quantity, hierarchical relationships, and attribute logic—but also constructs a quantifiable and traceable list of abnormal nodes. This provides a clear target scope for subsequent path reachability analysis and closed-loop evidence chain generation, thereby significantly improving the accuracy, robustness, and controllability of the entire configuration modeling process and effectively avoiding risks such as system deadlocks or call failures.

[0072] S300, for the list of nodes suspected of having topological loops, adopts the reachability cone domain traversal method to expand the vehicle configuration level from the root node, detects the topological fingerprint duplication during the traversal process and marks it immediately, and constructs a complete evidence chain of topological loop links.

[0073] In digital modeling, the presence of incorrect reference paths or structural backflow within the structure can easily lead to serious consequences such as recursive infinite loops, call failures, and ledger anomalies. To thoroughly identify and pinpoint such topological anomalies, after completing structural difference analysis and obtaining a list of nodes suspected of having topological loops, a reachability cone traversal operation starting from the root node of the vehicle configuration tree is required. This unfolds the structural path layer by layer, dynamically detects duplicate topological fingerprints, and immediately interrupts the unfolding upon discovering a closed-loop structure. Path details are recorded, and a closed-loop path chain is generated to construct a complete chain of evidence for topological loops. The specific operation process consists of the following steps:

[0074] Starting from the root node of the configuration tree structure, the starting point of the traversal is determined, and the path tracing process is initialized. The configuration root node refers to the highest-level structural unit corresponding to each vehicle in the structural representation, containing key information such as the vehicle's unique number in the operating platform, its vehicle model code, first deployment date, its route number, assigned maintenance unit information, and vehicle body structure version number. Before traversal, the latest structural state snapshot of the vehicle needs to be retrieved. This snapshot comes from the structural record table in the system's current configuration database and contains fields such as the complete path, topological fingerprint value, hierarchy depth, parent node identifier, number of child nodes, equipment function classification, equipment type name, node creation time, equipment status value, and deployment location number for each structural node. This snapshot serves as the data source for the traversal expansion, and it must be ensured that the recorded topological fingerprint and time anchor point are strictly consistent with the signature set generated in the previous stage. During traversal initialization, an empty path tracing linked list is also constructed to record the path structure, fingerprint information, and structural attributes traversed from the root node to the current structural node in real time, providing an accurate basis for subsequent fingerprint duplication detection.

[0075] During the traversal, all sub-structure nodes of the current node are progressively expanded in a depth-first manner. Depth-first means that for each structure node visited, all its direct sub-structure nodes are traversed first, then the next level of the substructure is recursively entered until all sub-paths have been traversed, at which point the parent node is backtracked to continue expanding other branches. During each structure expansion operation, the topological fingerprint of the current node is extracted from its structure path. The topological fingerprint is a unique structure identification value generated based on multiple fields such as structure path, structure location, device serial number, device parameter combination, level, and installation time. The algorithm for generating this value has been executed in previous steps. The topological fingerprint of the current node is compared one by one with the fingerprints of all historical nodes in the current path list. The comparison method is character-by-character matching. If a fingerprint value exactly the same as the current fingerprint is found in any historical path node, the path is immediately determined to have a closed-loop risk. At this point, the expansion of the current path is immediately interrupted, recursion is stopped, and the current path is marked as a "path with duplicate fingerprints". For example, if the path starts from "carriage number T1003 → third carriage → signal subsystem → signal main control equipment → signal subsystem", and the topological fingerprint of the second occurrence of "signal subsystem" is exactly the same as that of the first occurrence, then the path is determined to constitute a topological loop.

[0076] For all paths identified as having duplicate topological fingerprints, a complete closed-loop path chain is constructed as evidence of configuration anomalies. The evidence chain is constructed as follows: starting from the structural node where the duplicate fingerprint first appears, the information of the structural nodes recorded in the path chain is traced back level by level according to the original path order until the node where the duplicate fingerprint reappears is reached. In the entire backtracking path chain, the information of each structural node must be completely recorded, including but not limited to the following: node topological fingerprint value, structural path text, hierarchy depth value, device number, parent node path, child node path, creation time anchor point, device installation location number, device classification code, control type, associated communication address, acquisition port number, maintenance unit number, and device activation status. All this information is assembled into a closed-loop path chain in path order, forming a one-time complete structural reference trajectory. If the same structural path segment appears consecutively more than two nodes in the path chain, it must be marked as a high-risk closed-loop path. In scenarios where multiple closed-loop paths coexist, each path chain is stored independently, cannot be merged, and is numbered and classified according to the depth value, path length, and repetition frequency of the closed-loop path.

[0077] For all closed-loop path chains, path classification and structural tagging are performed for subsequent structural diagnosis, structural contraction, and risk scoring. Path classification is divided into three categories: First, direct closed-loop paths, where nodes reference themselves, and the path structure is characterized by the current node directly referencing its instance in the parent path; second, indirect closed-loop paths, where the current node returns to reference its ancestor node through several intermediate nodes, and this type of path may span multiple subsystems or device levels; and third, nested closed-loop paths, where the closed-loop path exists in the intersection of multiple parallel branches, and this structure often has a complex reference graph and is difficult to identify by static templates. After classification, a unique number is assigned to each closed-loop path chain, and the trigger node number, path entry node number, path depth, path length, duplicate fingerprint location, structural path text, and time anchor start and end values ​​are recorded. All closed-loop path chains are aggregated into a structural anomaly path database for use in topology problem tracing, data correction, structural reconstruction, version evolution diagram generation, and engineering verification.

[0078] This step aims to accurately identify structural loops, path backreferences, and hierarchical self-references in the configuration data by expanding all structural paths layer by layer from the root node of the vehicle configuration tree and comparing the topological fingerprint of each structural node in real time. Based on reachability cone traversal, this step does not rely on templates or manual judgment. Instead, it automatically captures potential structural errors leading to infinite recursion and system crashes by identifying fingerprint repetitions during the actual expansion of the configuration path. When a topological fingerprint repetition is detected, path expansion is immediately stopped, and the complete closed-loop structural path information is recorded backtracked according to the path expansion sequence, generating a chain of evidence for structural anomalies. These chains of evidence provide detailed location data for topological loop problems, which can be used not only for subsequent structural risk scoring and path repair operations but also to support dynamic verification, version tracking, and closed-loop control of configuration data. Through this step, the configuration modeling process acquires the ability to automatically diagnose and extract structural anomaly chains, significantly improving the stability, accuracy, and controllability of rail transit vehicle configuration data.

[0079] S400, based on the loop link topology, performs topology shrinkage on the loop substructure in an isolated sandbox environment, replaces the loop node with the structurally equivalent placeholder node, simulates the unfolding process and records the first position that cannot be terminated and the stack unfolding depth.

[0080] During digital modeling, after identifying topological loops in the configuration hierarchy, to verify whether the closed-loop path constitutes a recursion trap under actual operating conditions and to further quantify the impact of structural anomalies on stack behavior, a topological shrinkage operation needs to be performed on the closed-loop path in a simulation environment completely isolated from the production environment. Based on this, path simulation expansion is then performed to obtain the critical failure points and stack hierarchy limits in the structure expansion behavior. This step relies on the topological loop path chain extracted in the previous stage. By replacing key structural nodes and recording the process of abnormal expansion behavior, a measurable and traceable structural verification result is ultimately formed. The implementation consists of the following steps:

[0081] Extract the complete node sequence of each closed-loop path from the topological loop links constructed during the structural traversal phase, and identify the starting and ending nodes of the closed loop. The criterion for a closed loop is that two nodes in the structural path have completely identical topological fingerprints, and one node references the other node downwards on the hierarchical path, forming a backflow relationship. For example, in the structural path "Train No. T1009 → Second Carriage → Air Conditioning System → Control Unit → Air Conditioning System", the "Air Conditioning System" node appears twice, and its topological fingerprint is completely identical, indicating a path backflow. After confirming the closed loop, starting from the path where the node first appears, truncate downwards along the path to the repeating node, extracting the closed-loop structural segment as the target substructure for shrinkage processing. All structural nodes, path identifiers, fingerprints, hierarchical relationships, parent-child connection relationships, and device attributes in this structural segment are recorded and constructed as a structural segment copy for shrinkage replacement.

[0082] The target substructure described above replaces the original structural segment in the simulation environment as a structurally equivalent placeholder, and a logical mapping relationship is established between the placeholder structure and the original structure. The structurally equivalent placeholder node is created as follows: a new structural node entry is created, and the original structural segment's start path, end path, structural attribute summary, topology depth, number of connection points, equipment classification identifier, equipment parameter range, communication address range, and manufacturing information are embedded in the placeholder structure description in compressed form, while also being marked as a non-expandable structural node. This node is not resolved as a recursively expandable object during structural expansion; instead, it exists as an indivisible terminal in the structure tree, occupying its original position without triggering recursive behavior. After the replacement operation is completed, a compressed node will appear in the original structural path chain, maintaining the original connection between its parent node and the placeholder node, while the sub-path is broken. This replacement structure ensures structural integrity and path continuity, but eliminates the risk of circular references in the original structural segment, providing isolation control for simulation behavior.

[0083] After the structural replacement is completed, a simulated unfolding operation is performed on the entire path starting from the root node of the configuration to verify whether structural backflow still exists in the path when the replacement structure exists, or whether structural stack overflow occurs during the actual unfolding process. The execution order of the simulated unfolding operation strictly follows the original structural path. Whenever a structural node is visited, its path, topological fingerprint, structural number, and level information are pushed into the structural stack, and the current stack level is recorded. If the visited structural node is a placeholder structural node, the unfolding operation of that node is immediately terminated, and the structural unfolding process continues from the current node back to the previous expandable node. Throughout the simulation, each structural unfolding behavior is based on the path's forward direction. The structural stack expands as the path deepens and contracts as the path regresses. If abnormal duplicate structural fingerprints, non-convergent paths, or unfolding times exceed the maximum depth allowed by the configuration (e.g., setting the maximum level to 30 levels) occur during the simulation, the simulation is immediately interrupted, and the path, structural level value, structural fingerprint value, placeholder status, stack level, and memory usage of the node where the anomaly occurred are recorded. Taking the device node with the number "DJY-003" as an example, it appears twice in the path "Train No. T1009 → Fourth Carriage → Power System → Power Distribution Unit → Power Distribution Interface → Power Distribution Unit". If its occupancy status is not correctly identified during the simulation deployment, it will cause the stack to grow continuously until it exceeds the limit and is interrupted. This behavior is recorded as a failure of structure deployment.

[0084] The abnormal structural behaviors generated during the simulated deployment process are summarized and categorized to form a structural deployment failure report and a stack behavior analysis results table. This report must include the following key data fields: closed-loop path number, starting and ending node paths of the alternative structure, summary of placeholder structure attributes, location path where stack control first occurred, structural fingerprint, path depth, parent path identifier of the current node, maximum deployment stack depth, memory call levels at the time of deployment failure, successful rollback node paths, and failure trigger type (e.g., fingerprint re-entry, path breakage, placeholder identification failure). In addition, the report must also indicate the frequency of abnormal structural behaviors in the path, whether they affect other parallel paths, whether there are structural cross-references, and provide a structural snapshot view for each failed path. These results will serve as a direct basis for subsequent structural risk scoring and self-healing strategy execution, providing a precise and quantitative foundation for structural failure localization and path behavior evolution.

[0085] The purpose of this step is to verify the risk level of the closed-loop structure during actual deployment, based on the identified loop links in the topology, and to provide a quantitative basis for subsequent structural repair and control. By performing topology shrinkage on the structural path forming the closed loop in an isolated simulation environment, replacing the loop-forming node segments with structurally equivalent placeholder nodes, the backtracking path during the recursive deployment process can be effectively blocked. Based on this, the hierarchical structural deployment behavior from the root node downwards is simulated, dynamically recording the evolution of the structural stack and the location of the first abnormal node. This identifies path termination problems caused by incomplete loop processing, replacement failures, or deep nesting, and accurately calibrates the stack depth at that time. This mechanism not only comprehensively captures the substantial impact of loop paths on the integrity of the system configuration but also provides highly reliable operational empirical data for path risk level assessment, self-healing trigger judgment, and structural priority ranking, significantly improving the fault tolerance and engineering stability of urban rail transit vehicle configuration modeling.

[0086] S500 inputs the location where termination is not possible and the stack expansion depth as halting indicators into the verification engine, activates the cross-level structural consistency detection mechanism, and performs risk scoring on the reference structure between the bill of materials level, vehicle structure and vehicle equipment configuration.

[0087] In the process of urban rail transit vehicle configuration modeling, when topology expansion simulation identifies that some structural paths cannot complete normal recursion and the recorded structural expansion stack level exceeds the safety threshold, a structural consistency check and risk assessment mechanism needs to be triggered based on this abnormal behavior. This identifies potential systemic risks caused by issues such as path closure, reference conflicts, and structural drift. This operation uses "structural downtime indicators" as a starting point, integrating three dimensions: the bill of materials data table, the vehicle model structure standard template, and the vehicle equipment configuration list. Through path-by-path comparison, attribute difference quantification, hierarchical offset detection, and impact range assessment, it completes the scoring and determination of structural consistency and the archiving of high-risk nodes. This process is implemented in the following steps:

[0088] Each structural deployment failure event captured in the previous simulation step is used as an independent input to create a structural downtime index table. Each record contains at least the following fields: the complete path text of the structural deployment failure node, a unique topological fingerprint identifier, the vehicle number, the carriage number, the equipment system identifier, the stack depth at the time of the exception, the path identifier of the parent node at the location of the structural deployment interruption, the structural level number of the exception node, and the predefined level number of the node in the vehicle model structure standard template. This structural downtime index table needs to undergo a field integrity check to ensure that all data items originate from the path simulation data and equipment structure snapshots generated in the previous stage. Afterwards, all records are categorized by a combination of vehicle model number and vehicle number, and then sorted in ascending order by path level number to ensure that subsequent comparison operations can be performed sequentially and to avoid cross-reference errors.

[0089] Using each record in the structural downtime index table as a reference node, the structural information of that node in three data tables is extracted from the vehicle configuration database. The first data item comes from the "Vehicle Equipment Configuration List," which records the equipment deployment status under the current vehicle operating conditions, including the unique path of each structural node, equipment category code, equipment attribute parameters, hierarchical connection mapping, data acquisition port number, and actual installation time. The second data item comes from the "Vehicle Model Structure Standard Template Table," which defines the standard structural path hierarchy, structural node quantity limit, standard connection method, and parent-child dependency structure for each vehicle model during the theoretical design phase. For example, the deepest allowed path hierarchy for the air conditioning system under a certain vehicle model is 5 layers, and each air conditioning module is only allowed to be referenced by the controller within the same compartment and cannot be called across levels by the vehicle's main control system. The third data item is the "Vehicle Bill of Materials Hierarchy Table," which includes the part number, supply batch, assembly location, binding structural path, allowed connection hierarchy range, and supporting constraints for various components in the vehicle. These three tables constitute a consistent reference source for structural references. For each structural anomaly node, its corresponding fields are read from each table, and a horizontal field comparison relationship is established.

[0090] The data extracted from the three tables above are compared with the structural shutdown indicators to determine structural consistency. The comparison items include, but are not limited to, the following: 1) Structural path consistency: This determines whether the current node's path matches perfectly in the three tables. If there are missing fields, naming errors, or structural hierarchy shifts, the path is marked as inconsistent. 2) Equipment attribute differences: This compares the attribute field values ​​of the node in different tables, including equipment model, manufacturer code, control method, and data acquisition type. If any field has a content conflict, it is marked as inconsistent. 3) Parent-child reference stability: This checks whether the current node is within the allowed connection path range defined in the vehicle model template. If there are instances of lower-level nodes referencing higher-level nodes, cross-path references, or duplicate path calls, it is marked as a reference conflict. 4) Hierarchical number offset: This compares the node's hierarchical depth in the current vehicle configuration with the hierarchical number defined in the standard vehicle model template. If the offset exceeds the set tolerance (e.g., ±1 level), it is marked as structural drift. Each difference comparison generates a Boolean judgment value. If one or more inconsistencies exist, they constitute the scoring criteria.

[0091] All structural inconsistency results are converted into structural risk scores, and a structural risk score list is established. The risk score uses a 100-point scale, and the scoring formula is a weighted cumulative sum based on the number and severity of four types of discrepancies: path inconsistency is weighted at 30 points, attribute field conflict at 20 points, incorrect reference relationships at 30 points, and hierarchical drift at 20 points. If a node has both path inconsistency and hierarchical drift, its initial score is 30 + 20 = 50 points. If the node is also cross-referenced by multiple other paths, an additional 10 points are added for each cross-referenced path. The final risk score for each node is stored in the structural score list, including the node path, topological fingerprint, risk score, trigger timestamp, vehicle number, risk level label (e.g., high risk, medium risk, low risk), total number of referenced paths, total number of affected structural segments, and recommended handling method. The score list is sorted by risk level: nodes with a score higher than 75 are defined as high-risk nodes, those with a score between 50 and 74 are medium-risk nodes, and those with a score below 50 are low-risk nodes. All high-risk nodes need to be marked as mandatory targets for subsequent structural regulation, and a path reverse lookup channel needs to be established. This allows the indexing of all referenced paths, sources, and levels of reference for each node, providing a complete structural map for closed-loop governance.

[0092] The purpose of this step is to transform the abnormal shutdown locations and stack deployment depth detected during the structural deployment simulation into input criteria for structural consistency verification, comprehensively evaluating the correctness of reference and deployment stability of urban rail transit vehicle configuration data across multi-level structures. By comparing these shutdown indicators field-by-field with the paths, attributes, and connections in the vehicle bill of materials hierarchy table, vehicle model structure template table, and vehicle equipment configuration table, hidden problems such as structural path mismatches, equipment attribute conflicts, hierarchy number offsets, and parent-child relationship errors can be identified. Based on this, a structural risk score is calculated for each abnormal node according to the type and severity of structural differences, forming a risk list with hierarchy, directionality, and propagation impact. This scoring result not only provides a basis for subsequent structural repair and dynamic control but also serves as input conditions for configuration version verification, path tracing, and high-risk area early warning, thereby achieving dynamic supervision and hierarchical governance of vehicle structural data quality and improving the accuracy, controllability, and security of vehicle configuration management.

[0093] S600 activates the self-healing finite state control process based on the risk score results, performs a freeze operation on structural nodes that reach the risk critical threshold, and adjusts the structural introduction sequence through a fractional-order perturbation propulsion mechanism to achieve structural stability optimization of the vehicle configuration.

[0094] After completing the risk assessment of structural nodes in the urban rail transit vehicle configuration path, to prevent high-risk nodes from being incorrectly loaded or repeatedly referenced by path-dependent structures during configuration import, which could lead to data dead loops, non-terminating paths, or reference conflicts, a finite state control process needs to be introduced to dynamically freeze high-risk nodes and readjust their position order during the structure import process based on a perturbation mechanism. This control process follows deterministic state transition logic, combining freezing, perturbation, and import priority determination strategies to complete the processing of risk nodes through a stable closed-loop process. The process consists of the following steps:

[0095] From the previous stage's structural risk scoring list, all structural nodes with scores not lower than a preset freezing threshold (e.g., 75 points) are selected and included in the structural risk intervention list. Each node must have the following data fields: unique structural path identifier, topology fingerprint, vehicle number, vehicle model number, path depth number, maximum stack expansion level, risk score, number of referenced paths, list of parent structural paths, reference direction, attribute conflict type, and equipment category and number. For nodes with scores exceeding the freezing threshold, additional dynamic information such as their interruption position in the structural expansion simulation, the number of the loopback path they participated in, and whether they served as a recursion entry point must be recorded to form the basis for freezing operations. Subsequently, a structural freezing status mapping table is constructed, setting each high-risk node to a "frozen pending execution" state and indicating the freezing reason, freezing time limit, and reference isolation priority.

[0096] Before the structure import task is started, immediately perform a freeze operation on the aforementioned high-risk nodes. Freezing means removing the node from the current structure import path queue and prohibiting it from participating in any recursive expansion, path mapping, attribute binding, or device matching of structure topology paths. The freeze operation includes three aspects: First, blocking all direct reference paths to the node in all structure paths; second, writing the node's own attribute fields, installation anchor points, and path number into a structure placeholder node and marking it as "inactive"; third, adjusting the hierarchical offset in the original structure path and inserting a structure-neutral placeholder node to fill the gap and prevent path interruption. For example, if the node "Seventh Carriage - Brake Control Unit - XQZ-67" is marked as high-risk, its position in all vehicle configuration paths is frozen, and this position is replaced with a temporary structure node with a path placeholder code, device type label, and guiding attribute field. This structure node is not allowed to participate in path backtracking or be referenced by other devices to ensure that it does not have a diffuse impact on the system.

[0097] After the freezing operation is completed, to prevent the structure import process from being interrupted or the loading order of structure nodes from being disordered, it is necessary to perform an import order adjustment based on a perturbation-driven mechanism on the remaining unfrozen structure path queue. The basic logic of this mechanism is as follows: for all path segments that have direct or indirect structural reference relationships with the frozen nodes, perturbation sorting is performed according to their dependency strength, reference frequency, and path depth, so that high-stability paths are imported first and weakly dependent paths are processed later. The perturbation sorting operation includes the following steps: 1) Identify all upper-level path segments that reference the frozen nodes and calculate their structural dependency weight values. The weight values ​​are determined by the risk score of the referenced node, path depth, and reference level difference; 2) Reorder the path queue according to dependency weight from low to high; 3) Insert several structural buffer nodes, i.e., empty path nodes, into the original queue to reduce path dependency density and weaken loading impact intensity; 4) Redefine the path execution priority so that the placeholder node path is at the end, shared structural paths remain continuous, and low-risk paths are given priority; 5) Generate a new structure import order list. This mechanism effectively weakens the coupling tension of frozen nodes on the overall path structure during the import process and mitigates its potential loop impact, thereby providing preparation conditions for subsequent structural thawing and repair.

[0098] After the perturbation-ordered import process is ready, a simulation test is performed before structural loading to verify that all paths are continuous, without closed loops, and without broken links. The simulation test unfolds the structural nodes one by one according to the reordered path execution order and determines: 1) whether the path is interrupted; 2) whether a frozen node is invoked; 3) whether there are unreplaced placeholder nodes; and 4) whether there are path reverse references. If any of the above four checks fails, the perturbation order parameters are reset and the sorting process is re-executed until all paths meet the structural integrity requirements. After the structural import order stabilizes, it is written into the import execution plan list, and the recovery conditions for all frozen nodes are marked, including: risk score falling below the critical value, reference count reduced below the threshold, path depth repaired to template consistency, and attribute conflicts eliminated. Only after all conditions are met are frozen nodes allowed to restore their original structural paths and participate in structural reconstruction. The entire process forms a closed-loop control structure triggered by structural scoring, responded to by state freeze, mediated by disturbance sequence, and verified by import simulation, ensuring that the structural import behavior still has safety, controllability, and expected recoverability even when there are high-risk paths.

[0099] The purpose of this step is to proactively and dynamically respond to and adjust the import paths of high-risk structural nodes in the urban rail transit vehicle configuration based on their risk scores, ensuring the stability and controllability of the overall configuration structure during loading and application. Specifically, by setting a critical threshold for risk scores, high-risk nodes are temporarily frozen from the structure import sequence, preventing them from participating in the current configuration's path resolution and equipment attachment operations, thus preventing systemic structural anomalies caused by path closure, attribute conflicts, or reference disorder. Simultaneously, a disturbance propagation mechanism is introduced to reorder and adjust other structural paths affected by high-risk nodes, rationally inserting idle structural points and optimizing the import rhythm, thereby reducing reference coupling density and diffusion risks at the structural level. This state control and path optimization strategy, driven by risk scores, not only effectively isolates error propagation paths but also provides a stable environment for subsequent structural self-healing and unfreezing repair, ensuring that the vehicle configuration possesses high fault tolerance, recoverability, and engineering robustness in actual operation. This is a key step in achieving safe deployment of multi-level configuration modeling.

[0100] This invention generates unique topological fingerprints and constructs a hierarchical signature library, enabling each structural node to possess a verifiable identity and enhancing structural integrity verification capabilities. Simultaneously, it introduces a structural difference core and information entropy comparison mechanism, allowing potential loop nodes to be located and isolated during the data entry stage. Furthermore, by combining topological evidence chains with topological shrinkage experiments in a sandbox environment, it can simulate the real configuration loading process and capture the location of structural anomalies. Finally, through risk scoring and self-healing state control, it dynamically adjusts the structural import order, effectively breaking path dependency loops and achieving proactive structural stability and continuous optimization. Overall, this solution significantly improves the accuracy, controllability, and system robustness of rail transit vehicle configuration data and the operating platform, providing a solid data support foundation for digital operation and maintenance and intelligent scheduling across lines and vehicle types.

[0101] This invention provides, for example Figure 2 The illustrated digital modeling system for vehicle configuration in an urban rail transit system includes a hierarchical signature generation module, a structural difference kernel module, a loop evidence chain extraction module, a sandbox topology verification module, a risk scoring engine module, and a structural self-healing control module.

[0102] The hierarchical signature generation module establishes a global hierarchical baseline for vehicle configuration, generates a unique topological fingerprint for each node in the configuration hierarchy, and constructs a correspondence between the topological fingerprint and the time anchor point by combining the node creation time, forming a hierarchical signature library for comparing abnormal structures.

[0103] The structural difference branching kernel module executes the structural difference branching kernel based on the hierarchical signature library. It uses the quantity conservation rule, the hierarchical monotonic rule, and the parent-child information entropy threshold rule for parallel comparison to generate a list of nodes suspected of having topological loops.

[0104] The loop evidence chain extraction module targets a list of nodes suspected of having topological loops. It uses a reachability cone domain traversal method to expand the vehicle configuration level from the root node. During the traversal, it detects and immediately marks the duplicate topological fingerprints, thus constructing a complete evidence chain of loop links in the topological structure.

[0105] The sandbox topology verification module, based on the loop link topology structure, performs topology shrinkage on the loop substructure in an isolated sandbox environment, replaces the loop nodes with structurally equivalent placeholder nodes, simulates the unfolding process, and records the first position where it cannot be terminated and the stack unfolding depth.

[0106] The risk scoring engine module inputs the location where termination is impossible and the stack expansion depth as shutdown indicators into the verification engine, activates the cross-level structural consistency detection mechanism, and performs risk scoring on the reference structure between the bill of materials level, vehicle model structure and vehicle equipment configuration.

[0107] The structural self-healing control module activates the self-healing finite state control process based on the risk score results, performs a freeze operation on structural nodes that reach the risk critical threshold, and adjusts the structural import sequence through a fractional-order perturbation propulsion mechanism to achieve structural stability optimization of the vehicle configuration.

[0108] The present invention provides a digital modeling method for vehicle configuration of urban rail transit system, which is implemented by the above-mentioned digital modeling system for vehicle configuration of urban rail transit system. For details of the specific method and process of the digital modeling system for vehicle configuration of urban rail transit system, please refer to the embodiment of the above-mentioned digital modeling method for vehicle configuration of urban rail transit system, which will not be repeated here.

[0109] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for digital modeling of a configuration of a vehicle of an urban rail transit system, characterized in that, The method comprises the following steps: S100, establishing a global level baseline of the vehicle configuration, generating a unique topology fingerprint for each node in the configuration level, and combining the node creation time to build a correspondence between the topology fingerprint and the time anchor, forming a level signature library for abnormal structure comparison; S200, based on the level signature library, performing structure difference checking, using the quantity conservation rule, the level monotonicity rule and the parent-child information entropy threshold rule for parallel comparison, and generating a list of nodes suspected to have a topology loop; S300, for the list of nodes suspected to have a topology loop, expanding the vehicle configuration level from the root node using the reachability cone domain traversal method, detecting topology fingerprint repetition during the traversal process and immediately marking, and building a complete evidence chain of the topology structure loop link; S400, based on the topology structure loop link, performing topology contraction on the loop substructure in the isolation sandbox environment, replacing the loop node with a structure equivalent placeholder node, simulating the expansion process and recording the position and stack expansion depth where the process cannot be terminated for the first time; S500, inputting the position and stack expansion depth where the process cannot be terminated for the first time into the checking engine as a shutdown indicator, activating the cross-level structure consistency detection mechanism, and scoring the risk of the reference structure between the bill of materials level, the vehicle model structure and the vehicle device configuration; S600, according to the risk score result, activating the self-healing finite state control process, performing a freezing operation on the structure node that reaches the risk threshold, adjusting the structure import order through the fractional order disturbance promotion mechanism, and realizing the optimization of the structure stability of the vehicle configuration.

2. The method according to claim 1, wherein, Step S100 comprises: Collecting structure data for each type of urban rail transit vehicle, each vehicle, each car, each functional device and its subordinate components, and building a parent-child cascading structure path that gradually sinks from the whole vehicle level to the component level; Generating a topology fingerprint for each node in the structure path, which is generated by inputting the node full-path level block, structure attribute block and structure position block into the hash coding after being spliced in turn, to ensure its uniqueness in data dimension, time dimension and position dimension; Binding the topology fingerprint with the node creation time to build a one-to-one mapping relationship between the topology fingerprint and the time anchor, forming a structure time identifier; Arranging the topology fingerprints and time anchors of all structure nodes according to the vehicle number, car number, level path, fingerprint value and time stamp to build a full-level signature set covering all urban rail transit vehicle configuration structures.

3. The method according to claim 1, wherein, Step S200 comprises: Extracting the current structure snapshot of each urban rail transit train and comparing it with the manufacturing structure template at the path level to identify newly added nodes, missing nodes and path offset nodes; According to the comparison result, performing quantity conservation checking on each parent node to mark the first suspected abnormal node with abnormal number of child nodes or path return structure; Performing level monotonicity checking on the first suspected abnormal node to identify path level decrease, parent-child path overlap or path reverse structure, and recording the number of path overlap segments; Further performing parent-child information entropy difference analysis on the suspected node to calculate the difference between the attribute complexity of the parent node and the child node and the information entropy inversion coefficient, and forming a list of nodes suspected to have a topology loop.

4. The method according to claim 1, wherein, Step S300 includes: From the configuration tree root node of each urban rail transit vehicle, load the structure snapshot and initialize the path tracking linked list to ensure that the structure data is consistent with the hierarchical signature set; Expand the structure node layer by layer in a depth-first manner, and compare the current node topology fingerprint with the historical fingerprints in the path tracking linked list one by one in each expansion operation; If a fingerprint value identical to the current node topology fingerprint is found in the path, the current path expansion is immediately interrupted, and the repeated path information is recorded; Based on the repeated fingerprint path, a closed-loop path chain is constructed, and the topology fingerprint value, structure path, device number, parent-child relationship and time anchor point of all nodes in the path are recorded, and the closed-loop structure evidence chain is generated.

5. The method of claim 1, wherein, Step S400 includes: Extract the node sequence constituting the closed loop in the topology structure loop path chain, confirm the fingerprint repeated node and intercept the closed loop structure segment; Replace the closed loop structure segment with a structure equivalent placeholder node in the simulation environment, and establish the logical mapping relationship between the placeholder structure and the original structure; Based on the configuration root node, perform path simulation expansion operation, and record the position where the expansion cannot be terminated for the first time and the maximum expansion depth of the stack during the structure expansion; Classify and summarize the abnormal behaviors in the simulation expansion, and generate a structure expansion failure report containing the closed loop path number, fingerprint value, path depth and placeholder node state.

6. The method of claim 1, wherein, Step S500 includes: According to the structure shutdown index, a structure shutdown index table is established, and the path text, topology fingerprint, belonging number and stack level in each record are integrity checked; Call the vehicle device configuration list, vehicle structure standard template and vehicle material list level table, extract the field content related to the shutdown index and establish the horizontal comparison relationship; Based on the path consistency, device attribute difference, parent-child reference relationship stability and hierarchical number offset, the structure consistency is compared, and the structure difference judgment result is generated; Convert the structure difference result into a structure risk score value, aggregate to form a structure risk score list, and mark the structure nodes with a score higher than the set threshold as high-risk nodes.

7. The method of claim 1, wherein, Step S600 includes: Screen all structure nodes with a risk score not lower than the set threshold, generate a structure risk intervention list, and build a structure frozen state mapping table; Before structure import, perform freezing operation on all high-risk nodes in the structure risk intervention list, replace them with structure placeholder nodes, and block all direct references in the structure path; Perform import order adjustment operation based on the fractional order disturbance promotion mechanism on the remaining structure path queue, reorder the path loading order and insert the structure buffer node to reduce the path dependence coupling; According to the disturbance sorting result, perform structure loading simulation test to verify the structure path continuity and stability, complete the import execution plan and set the unfreezing recovery conditions of the frozen nodes.

8. A digital modeling system for a vehicle configuration of an urban rail transit system, for implementing a digital modeling method for a vehicle configuration of an urban rail transit system according to any one of claims 1 to 7, characterized in that, It includes a hierarchical signature generation module, a structure difference checking module, a loop evidence chain extraction module, a sandbox topology verification module, a risk score engine module and a structure self-healing regulation module: The hierarchical signature generation module establishes a global hierarchical baseline of the vehicle configuration, generates a unique topology fingerprint for each node in the configuration hierarchy, and builds a correspondence between the topology fingerprint and the time anchor based on the node creation time to form a hierarchical signature library for abnormal structure comparison. The structure difference checking module performs structure difference checking based on the hierarchical signature library, uses the quantity conservation rule, the hierarchical monotonicity rule, and the parent-child information entropy threshold rule for parallel comparison, and generates a list of nodes suspected of having a topology loop. The loop evidence chain extraction module uses the reachability cone domain traversal method to expand the vehicle configuration hierarchy from the root node, detects topology fingerprint duplication during the traversal process, and immediately marks it, and builds a complete evidence chain of the topology structure loop link. The sandbox topology verification module performs topology contraction on the loop substructure in an isolated sandbox environment based on the topology structure loop link, replaces the loop node with a structure equivalent placeholder node, simulates the expansion process, and records the position where termination fails for the first time and the stack expansion depth. The risk score engine module inputs the position where termination fails for the first time and the stack expansion depth into the checking engine as shutdown indicators, activates the cross-level structure consistency detection mechanism, and performs risk scoring on the reference structure between the bill of materials hierarchy, the vehicle model structure, and the vehicle device configuration. The structure self-healing regulation module activates the self-healing finite state control process based on the risk score results, performs a freeze operation on the structure nodes that reach the risk threshold, adjusts the structure import order through the fractional order disturbance propulsion mechanism, and optimizes the structural stability of the vehicle configuration.

Images