Wheel rail damage monitoring method based on damage tolerance design and digital twin drive

Abstract

The invention provides a wheeltrack damage monitoring method based on damage tolerance design and digital twinborn driving, and belongs to the technical field of wheeltrack damage monitoring, and the method comprises the steps: 1) constructing a wheeltrack system digital twinborn body module; 2) establishing a wheel track damage tolerance database; the method comprises the steps of (1) data feedback, (2) data feedback, (3) deployment of a multi-source sensing system, (4) damage state real-time mapping, (5) damage limit intelligent judgment, (6) a machine learning optimization module and (7) generation of a grading early warning instruction.The method can solve the problem that evaluation indexes fed back by existing twin data driven damage monitoring model data and existing damage tolerance design cannot be applied to the service process of a wheel-rail system in real time.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wheel-rail damage monitoring, in particular to a wheel-rail damage monitoring method based on damage tolerance design and digital twin driving. BACKGROUND

[0002] As one of the most core key components of rail transit, the service behavior of wheel-rail directly relates to the safe operation of trains. Once the wheel-rail fails due to fatigue and the corresponding repair measures are not taken in time, it will cause disastrous accidents of train destruction and human casualties. Therefore, in order to ensure the safe service performance of the wheel-rail system under complex environment, it is necessary and urgent to carry out intelligent monitoring and safe service evaluation of train wheel-rail damage.

[0003] At present, the research on the service safety of wheel-rail mainly focuses on the research of single macro factor, and basically does not involve the safety research of wheel-rail service damage, especially the service reliability evaluation research based on the damage tolerance of wheel-rail material. Therefore, it is necessary to carry out innovative research on the safe service evaluation technology of wheel-rail under complex environment considering the damage tolerance of material. At the same time, the digital twin technology is mainly used for the online monitoring of tool wear based on general wear model in the manufacturing process of lathe and other manufacturing industries, and the application of digital twin technology in the field of rail transit is mainly concentrated in the whole life cycle monitoring of rail transit equipment, and basically does not involve the dynamic monitoring and life prediction based on the service damage behavior of material wheel-rail. Therefore, it is necessary to construct a virtual digital twin based on the physical entity of wheel-rail, realize the virtual-real interaction of wheel-rail damage twin data, take the damage tolerance design value of wheel-rail as the damage state monitoring threshold, and clear the wheel-rail damage state early warning mechanism and the dynamic intelligent monitoring technology of wheel-rail service life under virtual-real interaction. This has important theoretical support and technical guidance for ensuring the safe operation and reliable service of high-speed railway wheel-rail under complex environment. SUMMARY

[0004] The present application provides a wheel-rail damage monitoring method based on damage tolerance design and digital twin driving, which can solve the problems of evaluation index of data feedback of existing twin data driven damage monitoring model and real-time application of existing damage tolerance design in wheel-rail system service process.

[0005] To achieve the above purpose, the present application adopts the following technical scheme: The wheel-rail damage monitoring method based on damage tolerance design and digital twin driving comprises: 1) constructing a wheel-rail system digital twin body module: integrating a physical sensor network, a data processing unit and a virtual twin model, for realizing the interactive mapping of wheel-rail physical entity and virtual twin model; 2) Establishing a wheel-rail damage tolerance database: storing initial damage tolerance thresholds determined based on wheel-rail material fatigue damage analysis research, and a residual life prediction model for calculating residual life based on the initial damage tolerance thresholds, the initial damage tolerance thresholds including critical crack size, crack propagation rate threshold and residual life prediction model; adjusting the initial damage tolerance thresholds based on real-time environmental parameters and material degradation model to obtain corrected dynamic damage tolerance thresholds, realizing dynamic damage tolerance correction, and updating the dynamic damage tolerance thresholds to the wheel-rail damage tolerance database; 3) Deploying a multi-source sensing system: deploying strain sensors, accelerometers, acoustic emission devices and damage topography monitoring units at key parts of the wheel-rail for collecting multi-source sensing data during wheel-rail operation; 4) Real-time mapping of damage state: processing the multi-source sensing data collected by the multi-source sensing system in step 3) through a cross-modal attention fusion network to extract wheel-rail damage features and generate real-time damage parameters that can represent the current damage condition of the wheel-rail; 5) Damage limit intelligent decision: using a crack parameter and damage topography dual criterion mechanism to compare the real-time damage parameters obtained in step 4) with the updated dynamic damage tolerance thresholds in the wheel-rail damage tolerance database in step 2), to obtain a comparison result containing damage overrun degree, and triggering an alarm according to the comparison result; 6) Machine learning optimization module: based on historical damage data and multi-source sensing data collected in step 3), the residual life prediction model and the cross-modal attention fusion network are optimized through self-supervised learning and virtual simulation training of the virtual twin model in step 1); 7) Generating a hierarchical warning instruction: generating a first, second and third level warning signal according to the damage overrun degree in the comparison result in step 5), and pushing the warning signal to the operation and maintenance terminal.

[0006] In this specification, the wheel-rail system digital twin module in step 1) includes the following three-layer structure: i) Physical layer: deploying a distributed sensor network and an edge computing node on the wheel-rail, the distributed sensor network being a specific implementation form of the physical sensor network, and the edge computing node being used for preliminary processing of data collected by the sensor; ii) Virtual layer: including a multi-scale modeling architecture, macro scale uses a wheel-rail finite element model to simulate crack propagation, meso scale uses a crystal plastic finite element model to simulate grain slip, and micro scale uses a molecular dynamics model to simulate dislocation evolution, and the micro-meso-macro damage correlation is established through the above multi-scale modeling; iii) Data layer: a spatio-temporal correlation database integrating real-time monitoring data stream and historical damage database, and a distributed machine learning inference engine deployed in the spatio-temporal correlation database, the inference engine being used for performing a damage state classification task in real time and working in cooperation with the data processing unit.

[0007] In the present specification, the wheel-rail damage tolerance database establishment process of step 2) comprises: i) Establishing a three-dimensional wheel-rail fatigue crack propagation model with an initial crack: pre-patching initial cracks of different lengths and angles in a full-size wheel-rail finite element model, obtaining the stress state of the wheel-rail fatigue crack tip after applying wheel-rail contact load; determining 15% of the current crack length as the crack propagation step to form a new wheel-rail crack morphology, repeating the crack propagation simulation until the crack length reaches the critical crack size, obtaining the function relationship between the cycle period and the crack length by polynomial fitting, and obtaining the function relationship between the crack length and the stress intensity factor range by polynomial fitting based on the fatigue crack propagation theory; extracting the stress parameters under different crack lengths, calculating the stress intensity factor and the reference stress at the crack tip, and then obtaining the function relationship between the crack length and the stress intensity factor, and the function relationship between the crack length and the reference stress by polynomial fitting; ii) Carrying out safety assessment of wheel-rail material with crack defects based on the stress state of the wheel-rail material fatigue crack tip: using the ratio of the stress intensity factor to the material fracture toughness to represent the fracture ratio of the wheel-rail material resisting fracture failure, and using the ratio of the reference stress to the flow stress to represent the load ratio of the wheel-rail material resisting plastic limit failure; based on the load ratio and the fracture ratio, carrying out the wheel-rail safety assessment based on the BS7910 standard, and using the first level assessment for the wheel-rail material; using the intersection of the function relationship between the crack length and the stress intensity factor, the function relationship between the crack length and the reference stress, and the standard failure assessment curve, the critical thermal crack size of the wheel is calculated, and further the residual life corresponding to the critical thermal crack size of the wheel is obtained by polynomial fitting of the crack length and the cycle number; iii) Preparing a damage tolerance design spectrum: combining a large number of wheel-rail fatigue contact test and simulation analysis results to prepare the damage tolerance design spectrum, in addition to crack data, the damage tolerance design spectrum also includes material wear amount and damage morphology as a damage tolerance judgment reference; iv) Establishing an environment-load-material performance dynamic coupling module: the module involves temperature effect correction parameters and load fluctuation correction parameters, based on the above correction parameters, dynamic critical crack size calculation is carried out to realize dynamic adjustment of the damage tolerance threshold, which matches the "adjusting the damage tolerance threshold based on real-time environmental parameters and material degradation model".

[0008] In the present specification, the multi-source sensing system in step 3) specifically comprises: i) Pressure sensor array: used for monitoring the wheel-rail contact stress distribution, providing stress data support for subsequent damage analysis; ii) Acoustic emission sensor: used for capturing the characteristic frequency in the process of wheel-rail crack propagation, identifying crack propagation behavior; iii) Embedded piezoelectric accelerometer: used for collecting vibration energy characteristics in the process of wheel-rail operation, reflecting the wheel-rail contact state; iv) Damage topography monitoring unit: including high-speed camera and laser confocal scanner, used for obtaining the topographic information of wheel-rail damage, providing topographic data for damage state evaluation.

[0009] In the specification, the cross-modal attention fusion network in step 4) includes the following parts: i) Modal feature extraction branch: 1D-CNN, 2D-CNN and ResNet-50 are used to process acoustic emission signals, vibration signals and topographic images in multi-source sensor data respectively, to extract damage features corresponding to each modality; ii) Cross-modal attention fusion module: through self-attention mechanism, the correlation weight between different modalities is learned, for example, the strong correlation between acoustic emission frequency and vibration energy during crack propagation, to realize effective fusion of multi-modal features; iii) Contrastive learning optimization: using NT-Xent loss function for training, to maximize the similarity of same type of damage features and maximize the difference of different types of damage features, to improve the recognition ability of the model for rare damage modes (such as abnormal wear); iv) Damage mode vector construction and risk coefficient calculation: based on the damage topographic features of wheel-rail materials, a damage mode vector is constructed, the topographic vector is fused with mechanical parameters and input into a virtual twin model, and a topographic risk coefficient is output, and a critical value of the topographic risk coefficient is determined according to engineering experience and related evaluation standards, to provide a basis for subsequent damage to limit decision.

[0010] In the specification, the damage to limit intelligent judgment double criterion mechanism in step 5) is specifically: i) Dynamic crack parameter criterion: when the crack size obtained by real-time monitoring is greater than the damage tolerance design size in the damage tolerance database, or the crack propagation rate is greater than the dynamic threshold in the database, it is determined that the criterion is met; ii) Topographic risk coefficient criterion: when the calculated damage topographic risk coefficient is greater than the preset critical value, it is determined that the criterion is met; iii) Early warning trigger rule: any of the above criteria is met, triggering an early warning, and if both criteria are met, the early warning level is raised to ensure that damage risks are not missed, consistent with the logic of "triggering an early warning".

[0011] In the specification, the machine learning optimization module of step 6) includes the following processes: i) Self-supervised pre-training process: By rotating the unlabelled vibration data for enhancement, the TCN network is trained to predict the signal rotation angle, and the damage data is enhanced; the bottom structure of the TCN network is frozen, and only the top network is fine-tuned for crack propagation direction prediction, so as to realize model pre-training based on normal wheel-rail operation data, and lay a foundation for subsequent model optimization; ii) Twin simulation-reinforcement learning closed loop: Simulate a large number of virtual damage scenarios in the virtual twin model, such as extreme load, material defect, etc., to generate virtual training data; use reinforcement learning to train the damage assessment model, and take "minimum prediction error" as the reward function to quickly iterate and optimize the model parameters in the virtual environment; iii) Model deployment: When the model prediction error is lower than the preset threshold, the optimized model is deployed to the corresponding physical system of the physical sensor network to realize the actual application of the model and complete the "virtual pre-training-physical fine-tuning" iterative optimization process.

[0012] In this specification, the hierarchical early warning instruction in step 7) is associated with the operation and maintenance strategy library, and the specific association relationship is: i) First-level (observation level) early warning: trigger the suggestion of rail grinding or wheel lapping, and start the fine-tuning process of the self-supervised model to optimize the monitoring model in time; ii) Second-level (maintenance level) early warning: generate a mandatory instruction for rail grinding or wheel lapping, and activate the virtual simulation training process to optimize the maintenance scheme based on virtual scenarios; iii) Third-level (emergency shutdown level) early warning: link the train control system to realize speed limiting operation, and trigger the reinforcement learning parameter optimization process to quickly improve the adaptability of the model to extreme damage scenarios; The associated actions of the operation and maintenance strategy library ensure that the early warning signal can be converted into actual operation and maintenance measures to realize the closed-loop management of "monitoring-early warning-operation and maintenance".

[0013] In this specification, the specific specifications of the initial cracks prepared in step 2) i) are: lengths of 0.5 mm, 1.0 mm, and 1.5 mm, and angles of 30°, 45°, and 60°; the applied wheel-rail contact load is simulated based on the Hertz contact theory under the condition of 21t axle load to ensure that the crack propagation simulation is consistent with the actual wheel-rail service conditions; the stress intensity factor is calculated by the M integral method, and the reference stress at the crack tip is calculated by the effective net section method, which provides an accurate parameter basis for subsequent function fitting and safety evaluation.

[0014] In the specification, the damage mode vector constructed in step 4) iv) specifically includes topographic feature parameters such as the area, depth, shape irregularity, etc. of the wheel-rail damage; the mechanical parameters include the contact stress monitored by the pressure sensor array, the stress intensity factor calculated by the stress, etc.; when the topographic vector and the mechanical parameters are fused, feature normalization processing is adopted to ensure the balanced weight of different dimension parameters; the determination of the topographic danger coefficient critical value needs to be combined with the fatigue performance test data of the wheel-rail material, historical damage failure cases and industry wheel-rail damage evaluation standards to ensure the rationality and reliability of the critical value, thereby providing an accurate basis for the damage-to-limit decision.

[0015] In summary, the present application has at least the following beneficial effects: The present application provides a wheel-rail damage monitoring method based on damage tolerance design and digital twin driving. Compared with the existing traditional fatigue analysis method based on nominal stress or crack-free life, the present application introduces the mature "damage tolerance" concept in the aviation, nuclear power and other fields into the field of wheel-rail fatigue analysis by considering the inevitable initial defects or small cracks in the wheel-rail system, and quantitatively calculates the probability and life of the expansion of these cracks under service load until the critical size. This is more in line with engineering practice, especially for wheel-rail materials subjected to high-frequency and high-stress alternating loads.

[0016] Meanwhile, compared with the existing wheel-rail damage monitoring and control measures, the twin data driven wheel-rail damage monitoring method not only can monitor the wheel-rail damage state in real time, but also can combine the damage tolerance design threshold to feedback the repair measures such as rail grinding and wheel turning to the wheel-rail service system in time, so as to realize "virtual-real interaction and virtual control of real". The wheel-rail damage intelligent monitoring new technology of damage tolerance and digital twin fusion will provide important theoretical and practical value for the safe and reliable service of wheel-rail. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] Figure 1 The schematic diagram of the wheel-rail damage monitoring method based on damage tolerance design and digital twin driving involved in the present application.

[0019] Figure 2 The schematic diagram of the wheel-rail rolling test bed digital twin involved in the present application.

[0020] Figure 3a The schematic diagram of the wheel-rail rolling test bed digital twin system involved in the present application.

[0021] Figure 3b Schematic diagram of twin data driven running process involved in the present application.

[0022] Figure 4a Schematic diagram of wheel-rail three-dimensional finite element model involved in the present application.

[0023] Figure 4b1 Schematic diagram of wheel-rail prefabricated crack (angle 30°) involved in the present application.

[0024] Figure 4b2 Schematic diagram of wheel-rail prefabricated crack (angle 45°) involved in the present application.

[0025] Figure 4b3 Schematic diagram of wheel-rail prefabricated crack (angle 60°) involved in the present application.

[0026] Figure 5a Schematic diagram of equivalent stress intensity factor with crack length relationship in crack propagation process (initial crack length is 0.5 mm) involved in the present application.

[0027] Figure 5b Schematic diagram of equivalent stress intensity factor with crack length relationship in crack propagation process (initial crack length is 1.0 mm) involved in the present application.

[0028] Figure 5c Schematic diagram of equivalent stress intensity factor with crack length relationship in crack propagation process (initial crack length is 1.5 mm) involved in the present application.

[0029] Figure 6a Schematic diagram of reference stress with crack propagation length change under 21t axle load (initial crack length is 0.5 mm) involved in the present application.

[0030] Figure 6b Schematic diagram of reference stress with crack propagation length change under 21t axle load (initial crack length is 1.0 mm) involved in the present application.

[0031] Figure 6c Schematic diagram of reference stress with crack propagation length change under 21t axle load (initial crack length is 1.5 mm) involved in the present application.

[0032] Figure 7a Schematic diagram of wheel-rail failure evaluation result of 21t axle load (initial crack length is 0.5 mm) involved in the present application.

[0033] Figure 7bA schematic diagram of the wheel-rail failure evaluation result (initial crack length of 1.0 mm) involved in the 21t axle load in the present application.

[0034] Figure 7c A schematic diagram of the wheel-rail failure evaluation result (initial crack length of 1.5 mm) involved in the 21t axle load in the present application.

[0035] Figure 8 A schematic diagram of the cross-modal attention fusion network involved in the present application.

[0036] Figure 9 A schematic diagram of the damage-to-limit intelligent judgment and grading early warning instruction flow involved in the present application.

[0037] Figure 10 A schematic diagram of the machine learning optimization module flow involved in the present application. DETAILED DESCRIPTION

[0038] In the following, only certain exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the embodiments of the present application. Therefore, the drawings and the description are considered to be essentially exemplary rather than limiting.

[0039] The following disclosure provides many different embodiments or examples for implementing different structures of the embodiments of the present application. In order to simplify the disclosure of the embodiments of the present application, the components and settings of specific examples are described in the following. Of course, they are only examples, and the purpose is not to limit the embodiments of the present application. In addition, the embodiments of the present application can refer to the same reference numerals and / or reference letters in different examples, and such repetition is for the purpose of simplification and clarity, which does not indicate the relationship between the various embodiments and / or settings discussed.

[0040] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0041] As Figure 1 shown, the present embodiment provides a wheel-rail damage monitoring method based on damage tolerance design and digital twin driving, which includes the following steps: 1) Building a wheel-rail system digital twin module: This module integrates a physical sensor network, a data processing unit, and a virtual twin model, which is used to realize the interactive mapping of the wheel-rail physical entity and the virtual twin model; wherein the virtual twin model can provide the virtual scene required for subsequent virtual simulation training, and the data processing unit can assist the preliminary operation of the subsequent damage monitoring related model, providing the basic architecture for the overall monitoring process; 2) Establish a wheel-rail damage tolerance database: first, store two types of core content - one is the initial damage tolerance threshold based on wheel-rail material fatigue damage analysis and research (the initial damage tolerance threshold is a quantitative indicator, including the critical crack size, crack propagation rate threshold), the second is the residual life prediction model based on the initial damage tolerance threshold to calculate the residual life (belongs to the damage monitoring related model, used to output the residual life combined with real-time damage parameters); then adjust the initial damage tolerance threshold based on real-time environmental parameters and material degradation model to obtain the corrected dynamic damage tolerance threshold, realize dynamic damage tolerance correction, and update the dynamic damage tolerance threshold and the updated residual life prediction model (based on dynamic threshold optimization calculation logic) to the wheel-rail damage tolerance database; 3) Deploy a multi-source sensing system: deploy strain sensors, accelerometers, acoustic emission devices and damage morphology monitoring units at key parts of the wheel and rail to collect multi-source sensing data during wheel and rail operation (this data is the "real-time monitoring data" mentioned in the subsequent steps, including stress, vibration, acoustic emission, damage morphology, etc. Dimensional information); 4) Real-time mapping of damage state: process the multi-source sensing data collected in step 3) through a cross-modal attention fusion network (belongs to the damage monitoring related model, used for multi-source data processing) to extract wheel and rail damage features and generate real-time damage parameters (including real-time crack size, real-time crack propagation rate, damage morphology feature parameters, etc.) that can represent the current damage situation; 5) Damage to limit intelligent decision: adopt a crack parameter and damage morphology dual criterion mechanism to compare the real-time damage parameters obtained in step 4) with the updated dynamic damage tolerance threshold in the database in step 2) to obtain a clear comparison result (the comparison result directly includes "damage whether exceeds limit" and "damage exceeds limit degree" information), and trigger the corresponding level of warning according to the comparison result; 6) Machine learning optimization module: based on historical damage data and multi-source sensing data (i.e. real-time monitoring data) collected in step 3), through self-supervised learning and virtual simulation training of the virtual twin model in step 1), optimize the residual life prediction model in step 2), the cross-modal attention fusion network in step 4) and other damage monitoring related models, and at the same time correct the adjustment logic of the dynamic damage tolerance threshold; 7) Generate graded warning instructions: generate a first (observation level), second (maintenance level), and third (emergency shutdown level) warning signal according to the damage over-limit degree in the comparison result in step 5), and push the warning signal to the operation and maintenance terminal to realize the direct association of warning and operation and maintenance requirements.

[0042] In some embodiments, the digital twin module construction in step 1) includes: i) physical layer: a distributed sensor network and edge computing nodes deployed on the wheel-rail; ii) virtual layer: including a multi-scale modeling architecture, macro-scale simulating crack propagation using a wheel-rail finite element model, meso-scale simulating grain slip using a crystal plasticity finite element model, micro-scale simulating dislocation evolution using a molecular dynamics model, establishing micro-meso-macro damage correlation; iii) data layer: a spatiotemporal correlation database integrating real-time monitoring data streams and historical damage databases, and deploying a distributed machine learning inference engine to perform real-time damage state classification tasks.

[0043] In some embodiments, the damage tolerance database establishment process of step 2) includes: i) there is an initial crack a 0 A three-dimensional wheel-rail fatigue crack propagation model is established. Different lengths and angles of initial cracks are pre-prepared in a full-size wheel-rail finite element model, and the stress state of the wheel-rail fatigue crack tip is obtained after applying wheel-rail contact load; 15% of the current crack length is determined as the crack propagation step to form a new wheel-rail crack morphology, and the crack propagation simulation is repeated until the crack length reaches a , the function relationship between the cycle period N and the crack length a is obtained by polynomial fitting, and the Paris formula ( , where C and n are material-related parameters) is used to obtain the function relationship between the crack length a and the stress intensity factor range ΔK ; stress parameters at different crack lengths are extracted, stress intensity factors are calculated by M-integral, reference stresses at the crack tip are calculated by effective net section method, and the function relationship between crack length and stress intensity factor and the function relationship between crack length and reference stress are obtained by polynomial fitting.

[0044] ii) Based on the stress state of the fatigue crack tip of the wheel-rail material, the safety evaluation of the wheel-rail material containing crack defects is carried out. The ratio of the stress intensity factor K to the material fracture toughness K IC characterizes the fracture ratio K r of the wheel-rail material resisting fracture failure; the ratio of the reference stress S ref to the rheological stress S f ( , σ Y is the yield strength, σ u is the tensile strength) characterizes the load ratio of the wheel-rail material resisting plastic limit failureL r ; based on load ratio L r and fracture ratio K r Based on BS7910 standard, the wheel-rail safety assessment is carried out, and the first level assessment is used for wheel-rail materials; by using the function relationship between crack length and stress intensity factor and the function relationship between crack length and reference stress, the intersection with the standard failure assessment curve is calculated to obtain the critical thermal crack size of the wheel, and further polynomial fitting of the crack length and the cycle number can obtain the residual life corresponding to the critical crack size of the wheel.

[0045] iii) A damage tolerance design spectrum is prepared by combining a large number of wheel-rail fatigue contact tests and simulation analysis results, and in addition to crack data, material wear amount and damage morphology should also be included as damage tolerance evaluation criteria.

[0046] First level assessment formula: ; : ; ; iv) A dynamic coupling module of environment-load-material performance is established, which involves temperature effect correction parameters K IC (T) and load fluctuation correction parameters β , based on which the dynamic critical crack size a c is calculated, and the dynamic adjustment of the damage tolerance threshold is realized.

[0047] K IC (T) = K IC0 · exp(- Q / RT); β =1+0.2(Δ P / P 0 ); a c = a c0 · K IC (T) / K IC0 · β ; In the formula, Q is the activation energy, R is the gas constant, and Δ P is the load fluctuation amplitude.

[0048] In some embodiments, the multi-sensor system in step 3) monitors the wheel-rail contact stress distribution using a pressure sensor array; captures the crack propagation characteristic frequency using an acoustic emission sensor; collects the vibration energy characteristics using an embedded piezoelectric accelerometer; and the damage morphology monitoring unit includes a high-speed camera, a laser confocal scanner, etc.

[0049] In some embodiments, the cross-modal attention fusion network in step 4) includes: i) a modal feature extraction branch: 1D-CNN, 2D-CNN, and ResNet-50 are used to process acoustic emission signals, vibration signals, and morphology images, respectively; ii) a cross-modal attention fusion module: the correlation weight between modalities is learned through a self-attention mechanism, such as the strong correlation between acoustic emission frequency and vibration energy during crack propagation; iii) contrastive learning optimization: using the NT-Xent loss function, the similarity of the same type of damage features is maximized, and the difference between the different types of damage features is maximized, improving the model's ability to recognize rare damage patterns (such as abnormal wear); iv) a damage pattern vector is constructed based on the wheel-rail material damage morphology characteristics, and the morphology vector and the mechanical parameters are fused into the twin body model, and the morphology risk coefficient is output ζ , and the critical value of the morphology risk coefficient is determined according to experience and related evaluation standards.

[0050] In some embodiments, the damage-to-limit decision in step 5) adopts a double-criteria mechanism: i) dynamic crack parameter criterion: the real-time crack size is greater than the damage tolerance design size or the crack propagation rate is greater than the dynamic threshold; ii) morphology risk coefficient criterion: the damage morphology risk coefficient ζ is greater than the critical value; iii) if any of the above criteria is met, an early warning is triggered, and if both criteria are met, the warning level is raised.

[0051] In some embodiments, the machine learning optimization module in step 6) includes: i) a self-supervised pre-training process: by rotating the unlabelled vibration data, the TCN network is trained to predict the signal rotation angle to achieve damage data enhancement, the bottom network is frozen to fine-tune the top layer for crack propagation direction prediction, and the model is pre-trained based on normal wheel-rail operation data; ii) a twin simulation-reinforcement learning closed loop: simulate a large number of virtual damage scenarios in the digital twin, such as extreme loads, material defects, etc., to generate virtual training data; use reinforcement learning to train the damage assessment model, and take "minimum prediction error" as the reward function to quickly iterate and optimize the model parameters in the virtual environment, and then deploy the optimized model to the physical system. Through this efficient iterative method of "virtual pre-training-physical fine-tuning", the model's adaptability to extreme working conditions is improved; iii) when the prediction error is below the threshold, the optimized model is deployed to the physical system.

[0052] In some embodiments, the hierarchical early warning instruction of step 7) is associated with an operation and maintenance strategy library: i) the first-level early warning triggers the rail grinding / wheel lapping suggestion, and starts the self-supervised model fine-tuning; ii) the second-level early warning generates the rail grinding / wheel lapping instruction, and activates the virtual simulation training; iii) the third-level early warning links the train control system speed limit, and triggers the reinforcement learning parameter optimization.

[0053] The technical concept of the present application is as follows: The wheel-rail damage monitoring method based on damage tolerance design and digital twin driving, as shown in Figure 1 includes the following steps: Step 1) Constructing a wheel-rail system digital twin module: integrating a physical sensor network, a data processing unit, and a virtual twin model, specifically involving: i) physical layer: deploying a distributed sensor network and edge computing nodes on the wheel-rail; ii) virtual layer: including a multi-scale modeling architecture, macro-scale using a wheel-rail finite element model to simulate crack propagation, meso-scale using a crystal plastic finite element model to simulate grain slip, micro-scale using a molecular dynamics model to simulate dislocation evolution, establishing micro-meso-macro damage correlation; iii) data layer: integrating a spatiotemporal correlation database of real-time monitoring data stream and historical damage database, and deploying a distributed machine learning inference engine to perform real-time damage state classification tasks.

[0054] Since the real wheel-rail system is in an open environment, there are many influencing factors such as environment, terrain, and operating conditions during its service process, and due to the difficulty and long cycle of field measurement tracking data, this embodiment considers starting with a commonly used laboratory scale double-disc rolling wheel-rail rolling contact simulation test machine, and establishes a wheel-rail rolling test bed digital twin as shown in Figure 2 , which involves the multi-source sensor arrangement in step 3) and the real-time mapping of damage state in step 4).

[0055] First, test data is collected and stored. A multi-type sensor system, including pressure, temperature, and torque sensors, is deployed on the MJP-30A wheel-rail rolling test platform to enable real-time monitoring of test parameters. The system utilizes a hierarchical data management architecture, enabling full data processing through the following methods: i) Dynamic Data Acquisition System: Sensors monitor test force, friction coefficient, slip ratio, and other motion parameters in real time. The data stream is stored in a dedicated local directory on the test platform in .txt format. Create corresponding data tables in the MySQL database in advance, covering static parameters such as sample material properties, and classified storage structures such as digital twin system calculation results; ii) Data transmission and storage: Dynamic data is transmitted across devices through the SMB network protocol. Within the same local area network, by configuring the IP address of the test equipment and shared folder permissions, a dedicated interface module is developed on the PC based on the Unity engine. The target folder is scanned according to the set cycle, and the updated content of the txt file is intelligently identified and parsed into structured data, which is finally classified and stored in the corresponding table of the database; iii) Static parameters are directly entered and stored in the specified data table through the Unity interactive interface; the digital twin data is generated by the wear calculation program and synchronously written to the database calculation layer through the API interface, forming a complete test data ecosystem.

[0056] Secondly, carry out twin data driven wheel-rail specimen wear prediction. Tγ / A For the wear model, the entire experimental process is first divided into intervals 1, 2, 3, ... according to the time column stored in the database. i 、 i +1, ..., n According to the real-time data recorded in the experiment, time= i The wear calculation and prediction are performed based on the test force, creep rate, friction coefficient, number of cycles of the wheel specimen, and the geometric parameters of the wheel-rail specimen, and the specimen radius is iteratively updated according to the wear threshold. Finally, a digital twin system of the wheel-rail rolling test bench wheel-rail specimen wear behavior was built in Unity. Figure 3a Based on this, the real-time mapping of the wear status of the wheel and rail materials (wear amount, wear depth, etc.) is achieved, and the twin data-driven operation process is as follows: Figure 3b shown.

[0057] Step 2) Establish a wheel-rail damage tolerance database: This database stores damage tolerance thresholds determined based on wheel-rail material fatigue damage analysis, including critical crack size, crack growth rate thresholds, and remaining life prediction models. This database adjusts the damage tolerance thresholds based on real-time environmental parameters and material degradation models to achieve dynamic damage tolerance correction. This embodiment conducts damage tolerance design based on the fatigue crack growth behavior of wheel-rail rolling contact, as follows: i) Establish the existence of initial cracks a 0three-dimensional wheel-rail fatigue crack propagation model. A three-dimensional wheel-rail finite element model (FEM) was established Figure 4a , and the stress state and cycle period N of the wheel-rail fatigue crack tip were obtained a ; initial cracks with different lengths (0.5 mm, 1.0 mm, 1.5 mm) and angles (30°, 45°, 60°) were implanted in the wheel-rail material, as shown in Figure 4b1 , Figure 4b2 and Figure 4b3 ; The rolling contact of the wheel on the rail under the condition of 21 t axle load was simulated based on the Hertz contact theory to obtain the stress state of the wheel-rail fatigue crack tip; the stress intensity factor was calculated based on the M-integral method according to the stress state of the wheel-rail fatigue crack tip, the wheel-rail crack propagation direction was determined based on the maximum energy release rate criterion, and the wheel-rail crack propagation rate was calculated using the Paris formula, which is , where C is 4.5966 x 10-13, n is 2.8805; 15% of the current crack length was determined as the crack propagation step, the new wheel-rail crack morphology was formed, and the stress state of the wheel-rail fatigue crack tip with the current length was obtained by implanting it into the original crack position and applying a load, and the crack propagation was repeated until the crack length reached a c The function relationship between the cycle period N and the crack length a was obtained by polynomial fitting, and the function relationship between the crack length a and the stress intensity factor range ΔK was obtained by polynomial fitting combined with the Paris formula; the stress parameters under different wheel-rail crack lengths were extracted, the stress intensity factor was calculated by the M-integral, and the reference stress at the crack tip was calculated by the effective net section method. The function relationship between the wheel-rail crack length and the stress intensity factor and the function relationship between the wheel-rail crack length and the reference stress were obtained by polynomial fitting, the relationship between the equivalent stress intensity factor and the crack length during the crack propagation process under the condition of 21 t axle load is as shown in Figure 5a (initial crack length 0.5 mm), Figure 5b (initial crack length 1.0 mm), and Figure 5c (initial crack length 1.5 mm), and the change of the reference stress with the crack propagation length under the condition of 21 t axle load is as shown in Figure 6a (initial crack length 0.5 mm), Figure 6b (initial crack length 1.0 mm), and Figure 6c (initial crack length 1.5 mm).

[0058] ii) Based on the stress state of the wheel-rail fatigue crack tip, the safety of the wheel-rail with crack defects was evaluated.

[0059] First, the stress intensity factor K is calculated The fracture ratio Rfcharacterizing the resistance of the wheel-rail to fracture failure is calculated K r , and the formula is ; (1) In the formula: K K is the stress intensity factor at the crack tip; K ⅠC Kc is the fracture toughness of the material.

[0060] The reference stress S ref The load ratio Rlcharacterizing the resistance of the wheel-rail to plastic limit failure is calculated L r , and the formula is ; (2) ; (3) In the formula: F P is the load borne by the structure, F Y Pc is the plastic limit load of the structure with a crack, S ref σ is the reference stress, S f σf is the flow stress, σ Y σy is the yield strength, σ u σt is the tensile strength.

[0061] Then, based on the load ratio Rl L r and the fracture ratio Rf K r The wheel-rail safety assessment based on BS7910 is carried out, and the first level assessment is adopted for the wheel-rail material (see formulas (4)-(7)), and the formula is ; (4) ; (5) ; (6) ; (7) In the formula: σy is the yield strength of the material, σt is the tensile strength, E E is the elastic modulus of the material. The wheel-rail failure evaluation results of the 21t axle load are as follows: Figure 7a (the initial crack length is 0.5mm), Figure 7b (the initial crack length is 1.0mm), and Figure 7c(initial crack length is 1.5 mm).

[0062] The cross-modal attention fusion network of step 4) is as shown in Figure 8 , including: i) a modal feature extraction branch: 1D-CNN, 2D-CNN and ResNet-50 are respectively used to process acoustic emission signals, vibration signals and topographic images; ii) a cross-modal attention fusion module: the correlation weight between modalities is learned through a self-attention mechanism, such as the strong correlation between acoustic emission frequency and vibration energy during crack propagation; iii) contrastive learning optimization: the NT-Xent loss function is used to maximize the similarity of the same damage features and maximize the difference of the different damage features, thereby improving the model's ability to recognize rare damage patterns (such as abnormal wear); iv) a damage mode vector is constructed based on the wheel-rail material damage topographic features, the topographic vector and the mechanical parameters are fused and input into the twin body model, and the topographic danger coefficient ζ is output, and the critical value of the topographic danger coefficient is determined according to experience and related evaluation standards.

[0063] The flow chart of step 5) damage to limit intelligent judgment and step 7) grading early warning instruction is as shown in Figure 9 : double-criterion judgment: a double-criterion mechanism of crack parameters and damage topography is used to compare real-time damage parameters with damage tolerance thresholds and trigger early warning, realizing the process as shown in Figure 8 , specifically: i) the crack size is greater than the damage tolerance design size; ii) the damage topographic danger coefficient ζ is greater than the critical value; iii) any of the above criteria is met to trigger early warning, and the early warning level is raised when both criteria are met.

[0064] Grading early warning instruction: according to the damage overrun degree, generate a first-level (observation level), second-level (maintenance level), and third-level (emergency shutdown level) early warning signal, and push it to the operation and maintenance terminal. The specific association operation and maintenance strategy library is: i) the first-level early warning triggers the steel rail grinding / wheel lapping suggestion, and starts the self-supervised model fine-tuning; ii) the second-level early warning generates a steel rail grinding / wheel lapping instruction, and activates the virtual simulation training; iii) the third-level early warning links the train control system speed limit, and triggers the reinforcement learning parameter optimization.

[0065] The flow chart of the machine learning optimization module of step 6) is as shown in Figure 10As shown, it includes: i) a self-supervised pre-training process: through rotation enhancement on unlabelled vibration data, a TCN network is trained to predict the signal rotation angle to realize the enhancement processing of damage data, the bottom network is frozen to fine-tune the top layer to predict the crack propagation direction, and the model pre-training based on normal wheel-rail operation data is realized; ii) a twin simulation-reinforcement learning closed loop: a large number of virtual damage scenarios such as extreme load, material defects, etc. are simulated in the digital twin, to generate virtual training data; the damage assessment model is trained by using reinforcement learning, and the "minimum prediction error" is taken as the reward function, the model parameters are quickly iterated and optimized in the virtual environment, and then the optimized model is deployed to the physical system. Through this efficient iterative way of "virtual pre-training-physical fine-tuning", the adaptability of the model to extreme working conditions is improved; iii) when the prediction error is lower than the threshold, the optimized model is deployed to the physical system.

[0066] The above-described embodiments are used to illustrate the present application, and are not intended to limit the present application, so the change of example values or the replacement of equivalent elements should still belong to the scope of the present application.

[0067] From the above detailed description, it is clear to those skilled in the art that the present application can achieve the above-mentioned purposes, and has met the requirements of the Patent Law.

[0068] Although the preferred embodiments of the present application have been described, those skilled in the art can make further changes and modifications to these embodiments once they know the basic inventive concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application. The above description is only the preferred embodiments of the present application and is not intended to limit the present application. It should be noted that any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

[0069] It should be noted that the above description of the process is only for example and illustration, and does not limit the scope of the present application. Those skilled in the art can make various modifications and changes to the process under the guidance of the present application. However, these modifications and changes are still within the scope of the present application.

[0070] The above has described the basic concept, and it is obvious that the above-mentioned invention disclosure is only as an example and does not constitute a limitation on the present application for those skilled in the art after reading this application. Although it is not explicitly stated here, those skilled in the art can make various modifications, improvements and modifications to the present application. Such modifications, improvements and modifications are suggested in the present application, so such modifications, improvements and modifications still belong to the spirit and scope of the exemplary embodiments of the present application.

[0071] At the same time, this application uses specific terms to describe the embodiments of this application. For example, "one embodiment," "an embodiment," and / or "some embodiments" refer to a certain feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that "one embodiment," "an embodiment," or "an alternative embodiment" mentioned twice or more in different places in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this application may be appropriately combined.

[0072] Furthermore, those skilled in the art will appreciate that various aspects of the present application may be illustrated and described in terms of a number of patentable categories or situations, including any new and useful process, machine, product, or combination of substances, or any new and useful improvement thereof. Thus, various aspects of the present application may be implemented entirely in hardware, entirely in software (including firmware, resident software, microcode, etc.), or in a combination of hardware and software. Each of the above hardware and software may be referred to as a "unit," "module," or "system." Furthermore, various aspects of the present application may take the form of a computer program product embodied in one or more computer-readable media, with computer-readable program code embodied therein.

[0073] The computer program code required for the operation of each part of this application can be written in any one or more programming languages, including object-oriented programming languages ​​such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc., conventional procedural programming languages ​​such as C programming language, Visual Basic, Fortran2103, Perl, COBOL2102, PHP, ABAP, dynamic programming languages ​​such as Python, Ruby and Groovy, or other programming languages. The program code can be run entirely on the user's computer, or as a standalone software package on the user's computer, or partly on the user's computer and partly on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer via any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service such as software as a service (SaaS).

[0074] Furthermore, the order of processing elements or sequences, or the use or appearance of certain terminology, throughout the above description should not be construed as limiting the application. Other steps, components, or configurations can be determined and implemented in a manner most beneficial to a particular application. For example, although the implementation of the various components described above can be embodied in hardware devices, it can also be implemented as a pure software solution, for example, as an installation on an existing server or mobile device.

[0075] Similarly, it is to be noticed that the term "comprising", used in the description, should not be interpreted as being restricted to the means listed thereafter; it does not exclude other elements or steps. It is thus to be interpreted as specifying the presence of the stated features, integers, steps or components as referred to, but does not preclude the presence or addition of one or more other features, integers, steps or components, or groups thereof. Furthermore, the description of the application is not intended to limit the application to the form disclosed herein. Various modifications and changes can be made without departing from the spirit and scope of the application as set forth in the following claims.

Claims

1. A wheel-rail damage monitoring method based on damage tolerance design and digital twin drive is characterized by: include: 1) Constructing a digital twin module for the wheel-rail system: Integrating a physical sensor network, a data processing unit, and a virtual twin model to achieve interactive mapping between the wheel-rail physical entity and the virtual twin model; 2) Establishing a wheel-rail damage tolerance database: storing the initial damage tolerance threshold determined based on wheel-rail material fatigue damage analysis, and the remaining life prediction model that calculates the remaining life based on the initial damage tolerance threshold. The initial damage tolerance threshold includes a critical crack size, a crack growth rate threshold, and a remaining life prediction model; the initial damage tolerance threshold is adjusted based on real-time environmental parameters and a material degradation model to obtain a corrected dynamic damage tolerance threshold, thereby implementing dynamic damage tolerance correction, and the dynamic damage tolerance threshold is updated in the wheel-rail damage tolerance database; 3) Deployment of a multi-source sensing system: strain sensors, accelerometers, acoustic emission devices, and damage morphology monitoring units are deployed at key locations on the wheel and rail to collect multi-source sensing data during wheel and rail operation; 4) Real-time damage status mapping: The multi-source sensor data collected by the multi-source sensor system in step 3) is processed through a cross-modal attention fusion network to extract wheel-rail damage characteristics and generate real-time damage parameters that can characterize the current damage status of the wheel and rail; 5) Intelligent judgment of damage limit: Using a dual-criteria mechanism of crack parameters and damage morphology, the real-time damage parameters obtained in step 4) are compared with the updated dynamic damage tolerance threshold in the wheel-rail damage tolerance database in step 2) to obtain a comparison result including the degree of damage exceeding the limit. An early warning is triggered based on the comparison result; 6) Machine Learning Optimization Module: Based on historical damage data and the multi-source sensor data collected in step 3), the remaining life prediction model and cross-modal attention fusion network are optimized through self-supervised learning and virtual simulation training of the virtual twin model in step 1). 7) Generate graded warning instructions: Generate level 1, level 2, and level 3 warning signals based on the damage exceeding the limit in the comparison result of step 5), and push the warning signals to the operation and maintenance terminal.

2. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 1 is characterized in that: The wheel-rail system digital twin module construction in step 1) includes the following three layers: i) Physical layer: Distributed sensor networks and edge computing nodes deployed on wheels and rails. The distributed sensor network is a specific implementation of a physical sensor network, and the edge computing nodes are used to perform preliminary processing on the data collected by the sensors. ii) Virtual layer: This includes a multi-scale modeling framework that uses a wheel-rail finite element model to simulate crack propagation at the macroscale, a crystal plasticity finite element model to simulate grain slip at the mesoscale, and a molecular dynamics model to simulate dislocation evolution at the microscale. This multi-scale modeling establishes a micro-meso-macro damage correlation. iii) Data layer: A spatiotemporal correlation database that integrates real-time monitoring data streams and historical damage databases is deployed, and a distributed machine learning inference engine is deployed in the spatiotemporal correlation database. The inference engine is used to perform damage status classification tasks in real time and work in collaboration with the data processing unit.

3. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 1 is characterized in that: The process of establishing the wheel-rail damage tolerance database in step 2) includes: i) Establishing a three-dimensional wheel-rail fatigue crack growth model with an initial crack: Initial cracks of different lengths and angles are prefabricated in a full-scale wheel-rail finite element model. After applying a wheel-rail contact load, the stress state at the wheel-rail fatigue crack tip is obtained. A crack growth step length of 15% of the current crack length is determined to form a new wheel-rail crack morphology. The crack growth simulation is repeated until the crack length reaches the critical crack size. The functional relationship between the cycle period and the crack length is obtained through polynomial fitting. In combination with relevant fatigue crack growth theories, the functional relationship between the crack length and the stress intensity factor range is obtained through polynomial fitting. Stress parameters at different crack lengths are extracted, and the stress intensity factor and reference stress at the crack tip are calculated. Finally, the functional relationship between the crack length and the stress intensity factor, as well as the functional relationship between the crack length and the reference stress, is obtained through polynomial fitting. ii) Safety assessment of wheel and rail materials containing crack defects is conducted based on the stress state at the fatigue crack tip of the wheel and rail material: the ratio of the stress intensity factor to the material's fracture toughness is used to characterize the fracture ratio of the wheel and rail material's resistance to fracture failure, and the ratio of the reference stress to the flow stress is used to characterize the load ratio of the wheel and rail material's resistance to plastic limit failure. Wheel and rail safety assessment is conducted based on the load ratio and fracture ratio, using a first-level assessment for the wheel and rail material. The critical thermal crack size of the wheel is calculated by intersecting the functional relationships between the crack length and the stress intensity factor, and the crack length and the reference stress, with the standard failure assessment curve. A polynomial fit is then performed between the crack length and the number of cycles to determine the remaining life corresponding to the critical crack size of the wheel. iii) Compiling a damage tolerance design spectrum: The damage tolerance design spectrum is compiled based on the wheel-rail fatigue contact test and simulation analysis results. In addition to crack data, the damage tolerance design spectrum also includes material wear and damage morphology as a criterion for judging damage limit; iv) Establish an environment-load-material performance dynamic coupling module: This module involves temperature effect correction parameters and load fluctuation correction parameters. Based on these correction parameters, dynamic critical crack size calculation is carried out to achieve dynamic adjustment of the damage tolerance threshold, which matches the adjustment of the damage tolerance threshold based on real-time environmental parameters and material degradation models.

4. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 1 is characterized in that: The multi-source sensing system in step 3) specifically includes: i) Pressure sensor array: used to monitor the wheel-rail contact stress distribution and provide stress data support for subsequent damage analysis; ii) Acoustic emission sensor: used to capture the characteristic frequencies during the wheel-rail crack propagation process and identify the crack growth behavior; iii) Embedded piezoelectric accelerometer: used to collect vibration energy characteristics during wheel-rail operation and reflect the wheel-rail contact status; iv) Damage morphology monitoring unit: includes a high-speed camera and a laser confocal scanner, used to obtain the morphological information of wheel-rail damage and provide morphological data for damage status assessment.

5. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 1 is characterized in that: The cross-modal attention fusion network in step 4) consists of the following parts: i) Modal feature extraction branch: 1D-CNN, 2D-CNN, and ResNet-50 are used to process the acoustic emission signals, vibration signals, and topographic images in the multi-source sensor data to extract the damage features corresponding to each modality; ii) Cross-modal attention fusion module: This module learns the association weights between different modalities through the self-attention mechanism to achieve effective fusion of multimodal features; iii) Contrastive learning optimization: Using the NT-Xent loss function for training, we maximize the similarity of similar damage features and the difference of heterogeneous damage features, improving the model's ability to recognize rare damage patterns. iv) Damage pattern vector construction and hazard factor calculation: A damage pattern vector is constructed based on the damage morphology characteristics of the wheel-rail material. The damage pattern vector is integrated with the mechanical parameters and input into the virtual twin model. The damage morphology hazard factor is output and the critical value of the damage morphology hazard factor is determined, providing a basis for subsequent damage limit judgment.

6. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 5 is characterized in that: The dual-criteria mechanism for damage-to-limit intelligent decision in step 5) is as follows: i) Dynamic crack parameter criterion: This criterion is considered to be met when the crack size obtained through real-time monitoring is larger than the damage tolerance design size in the damage tolerance database, or when the crack growth rate is larger than the dynamic threshold in the database; ii) Appearance hazard coefficient criterion: When the calculated damage appearance hazard coefficient is greater than the preset critical value, the criterion is determined to be met; iii) Warning triggering rules: Meeting any of the above criteria triggers a warning. If both criteria are met, the warning level is raised to ensure that no damage risk is missed.

7. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 1 is characterized in that: The machine learning optimization module in step 6) includes the following processes: i) Self-supervised pre-training process: By performing rotation enhancement on unlabeled vibration data, the TCN network is trained to predict the signal rotation angle, thereby enhancing damage data. The underlying structure of the TCN network is frozen, and only the top-level network is fine-tuned for crack propagation direction prediction, thus achieving model pre-training based on normal wheel-rail operation data. ii) Twin Simulation-Reinforcement Learning Closed Loop: Virtual damage scenarios are simulated in the virtual twin model to generate virtual training data. The damage assessment model is trained using reinforcement learning, minimizing prediction error as the reward function, and model parameters are rapidly iterated and optimized in the virtual environment. iii) Model deployment: When the model prediction error is lower than the preset threshold, the optimized model is deployed to the physical system corresponding to the physical sensor network to realize the practical application of the model and complete the iterative optimization process of virtual pre-training and physical fine-tuning.

8. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 7 is characterized in that: The hierarchical warning instructions in step 7) are associated with the operation and maintenance strategy library. The specific association relationship is: i) Level 1 Warning: This triggers a recommendation for rail grinding or wheel repair and initiates the fine-tuning process of the self-supervision model to optimize the monitoring model in a timely manner. ii) Level 2 Warning: Generates mandatory instructions for rail grinding or wheel repair, while activating a virtual simulation training process to optimize maintenance plans based on virtual scenarios. iii) Level 3 Warning: This system links with the train control system to implement speed limits and simultaneously triggers a reinforcement learning parameter optimization process to rapidly improve the model's adaptability to extreme damage scenarios. The associated actions of the operation and maintenance strategy library ensure that the early warning signals can be converted into actual operation and maintenance measures, realizing the closed-loop management of monitoring-early warning-operation and maintenance.

9. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 3 is characterized in that: The specific specifications of the initial cracks are: lengths of 0.5mm, 1.0mm, and 1.5mm, and angles of 30°, 45°, and 60°, respectively. The applied wheel-rail contact load is based on Hertz contact theory to simulate a 21t axle load condition, ensuring that the crack propagation simulation is consistent with actual wheel-rail service conditions. The M-integral method is used to calculate the stress intensity factor, and the effective net section method is used to calculate the reference stress at the crack tip.

10. The wheel-rail damage monitoring method based on damage tolerance design and digital twin drive according to claim 5, characterized in that: The damage pattern vector specifically includes the area, depth, and shape irregularity of the wheel-rail damage; the mechanical parameters include the contact stress monitored by the pressure sensor array and the stress intensity factor obtained through stress calculation; when fusing the morphology vector with the mechanical parameters, feature normalization is used to ensure balanced weights of parameters of different dimensions; the determination of the critical value of the damage morphology hazard factor requires a combination of fatigue performance test data of the wheel-rail material, historical damage failure cases, and industry wheel-rail damage assessment standards.

Images