Rail-wheel force active feedback type vibration absorbing system of track rail-wear section

Abstract

The application discloses a wheel-rail force active feedback type vibration absorption system of a track rail corrugation section and relates to the technical field of track transportation intelligent monitoring, and aims to solve the technical problem that the prior art cannot utilize the operation train self-control system operation data to perform high-frequency online dynamic evaluation on the outer rail side wear state without adding special sensors, and the wheel-rail force active feedback type vibration absorption system comprises the following modules: an acquisition module, which is used for acquiring the feedback force time sequence output by the wheel-rail force active feedback type vibration absorption system corresponding to the outer rail side control channel in real time when a train passes through a target curve section; and a transient detection module, which is used for detecting the feedback force time sequence and intercepting a first transient response waveform generated in the process that the train enters a circular curve from a transition curve. The application has the advantages that the asymmetry of the feedback force transient response when the vibration absorption system enters and exits the curve is used as a side wear criterion, and track disease detection and active vibration suppression share the same hardware platform.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring technology for rail transit, and more specifically, to an active feedback vibration absorption system for wheel-rail forces in track corrugation sections. Background Technology

[0002] In railway operations, lateral wear of the outer rail in curved sections is a common type of rail damage. With the accumulation of total weight, the material on the inner side of the outer rail head gradually peels away, forming an asymmetrical profile. This profile degradation worsens the wheel-rail contact geometry, increases lateral force and wear rate, and in severe cases, affects driving safety and passenger comfort. Therefore, timely and accurate dynamic assessment of the outer rail lateral wear condition is a crucial prerequisite for guiding rail grinding and maintenance and ensuring track smoothness.

[0003] Currently, the main methods for obtaining data on outer rail side wear include manual measurement using specialized calipers at fixed points and offline or on-board inspection using rail profile laser scanners. While these methods can provide relatively accurate wear values, they generally require independent measuring equipment, occupy track windows during inspections, and are difficult to implement for every trip, high-frequency online dynamic assessment on operating trains. On the other hand, existing technologies have developed methods to identify track corrugation and other defects using axle box acceleration or wheel-rail force signals, but these methods focus on periodic irregularities along the longitudinal direction of the rail, rather than the asymmetric degradation of the outer rail head profile in the transverse direction. The specific mechanical characteristics caused by outer rail side wear, namely the directional migration of the wheel-rail contact point along the wear slope when the train enters or exits a curve, have not been used as a direct source of information for assessing the side wear condition in traditional inspection approaches. Therefore, how to achieve online dynamic assessment of the outer rail side wear condition of curves without adding specialized sensors, using existing vehicle system operating data, has become an urgent technical problem to be solved. In view of this, we propose a wheel-rail force active feedback vibration absorption system for track corrugation sections. Summary of the Invention

[0004] The purpose of this invention is to provide an active feedback vibration absorption system for wheel-rail forces in track corrugation sections, in order to solve the technical problem that existing technologies cannot perform high-frequency online dynamic assessment of the external rail side wear status using the operating data of the train's own control system without adding dedicated sensors.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an active feedback vibration absorption system for wheel-rail forces in track corrugation sections, comprising: The acquisition module is used to acquire the time series of feedback force output by the wheel-rail force active feedback vibration absorption system corresponding to the outer rail side control channel in real time when the train passes through the target curve section; The transient detection module is used to detect the feedback force time series and extract the first transient response waveform generated when the train enters the circular curve from the transition curve, and the second transient response waveform generated when the train enters the transition curve from the circular curve. The feature extraction module is used to extract the overshoot features of the first transient response waveform and the second transient response waveform respectively, and to calculate the ratio of the extracted overshoot features to obtain the asymmetry ratio. The asymmetry ratio is used to characterize the degree of difference in the transient response of the feedback force when entering the curve and leaving the curve due to the asymmetry of the outer rail head profile. The side wear assessment module is used to dynamically assess the side wear status of the outer rail in the target curve section based on the pre-established mapping relationship between the asymmetry ratio and the wear status of the outer rail side.

[0006] Preferably, the acquisition module includes a mileage acquisition unit, a route data unit, and a trigger recording unit; The mileage acquisition unit is used to acquire the real-time mileage information of the train, the line data unit is used to store the start and end mileage data of the curve section of the target line, and the trigger recording unit is used to trigger the recording of the feedback force time sequence of the outer track side control channel when the real-time mileage information indicates that the train is about to enter the target curve section. The transient detection module includes a gradient calculation unit, a step recognition unit, and a waveform truncation unit; The gradient calculation unit is used to calculate the time gradient value of the feedback force time series. The step identification unit is used to determine that the current time is the start time of the transient event when the time gradient value exceeds the preset step identification condition within multiple consecutive sampling points. The waveform truncation unit is used to extract the corresponding transient response waveform by extending a preset time length forward and backward based on the start time. The corresponding transient response waveform is the one that matches the detected transient event type between the first transient response waveform and the second transient response waveform.

[0007] Preferably, the transient detection module further includes a validity verification unit; The validity verification unit is used to obtain bogie angle change information when the train passes through the transition curve, and compare the start time of the transient event with the transition curve passing time indicated by the bogie angle change information to confirm that the start time is within the transition curve passing time, so as to confirm that the captured transient response waveform corresponds to the mechanical disturbance stage of the train entering the circular curve and leaving the circular curve.

[0008] Preferably, the first transient response waveform corresponds to the response in the feedback force control output of the step disturbance caused by the wear slope on the inner side of the outer rail head causing the wheel-rail contact point to slip inward when the train enters the circular curve from the transition curve; The second transient response waveform corresponds to the response of the step disturbance generated by the outward rise of the wheel-rail contact point when the train enters the transition curve from the circular curve.

[0009] Preferably, the overshoot characteristics are overshoot peak value, overshoot decay time constant, and overshoot waveform energy asymmetry; Wherein, the overshoot peak value is the maximum amplitude in the transient response waveform that deviates from the steady-state value; The overshoot decay time constant is a time constant obtained by envelope fitting of the decay segment of the transient response waveform; The overshoot waveform energy asymmetry is the ratio between the positive energy exceeding the steady-state value and the negative energy below the steady-state value in the transient response waveform.

[0010] Preferably, the feature extraction module includes a ratio calculation unit and a fusion calculation unit; The ratio calculation unit is used to calculate the ratio of the first transient response waveform and the second transient response waveform under the same type of overshoot feature; The fusion calculation unit is used to weight and fuse the ratios corresponding to various overshoot features when the extracted overshoot features include at least two types to generate a comprehensive asymmetry index. The comprehensive asymmetry index is then input as an asymmetry ratio to the side wear assessment module. The ratio calculation unit calculates the ratio of the first transient response waveform and the second transient response waveform under the same type of overshoot feature in the following manner: When the overshoot characteristic is the overshoot peak value, let the overshoot peak value of the first transient response waveform be... The overshoot peak value of the second transient response waveform is Then the overshoot peak ratio satisfy: ; When the overshoot characteristic is the overshoot decay time constant, let the overshoot decay time constant of the first transient response waveform be... The overshoot decay time constant of the second transient response waveform is Then the decay time constant is greater than satisfy: ; When the overshoot characteristic is the energy asymmetry of the overshoot waveform, let the energy asymmetry of the first transient response waveform be... The energy asymmetry of the second transient response waveform is The energy asymmetry is compared to satisfy: ; The fusion computing unit generates the comprehensive asymmetry index in the following manner: When overshoot features include When planting, Let the first The ratio corresponding to the overshoot feature is The corresponding weighting coefficient is Then the comprehensive asymmetry index satisfy: ; in, To The normalized mapping function performed.

[0011] Preferably, the feature extraction module further includes a baseline compensation unit; The baseline compensation unit is used to obtain the baseline asymmetry ratio of the same target curve section under the standard track profile when multiple trains pass through, and to use the baseline asymmetry ratio to compensate the current calculation result when calculating the asymmetry ratio, so as to eliminate the background influence caused by curve superelevation and wheel load transfer. Wherein, the uncompensated asymmetry ratio output by the feature extraction module is denoted as The baseline asymmetry ratio is Then the asymmetry ratio after compensation satisfy: ; The side wear assessment module will It is used as an asymmetric ratio for dynamic evaluation.

[0012] Preferably, the side wear evaluation module includes a mapping model storage unit and an evaluation output unit; The mapping model storage unit is used to store pre-built mapping models. The mapping model defines the correspondence between the asymmetry ratio and the wear amount on the outer rail side. The mapping model is constructed according to the curve radius, rail material and total weight. The evaluation output unit is used to call the corresponding mapping model according to the actual line parameters of the target curve section, and input the asymmetry ratio into the called mapping model to output the outer rail side wear of the target curve section, and compare the outer rail side wear with the preset maintenance intervention level threshold to output the evaluation result of the outer rail side wear state of the target curve section. Wherein, let the asymmetry ratio after compensation be ? The mapping model called is The wear amount on the outer rail side satisfy: .

[0013] Preferably, it also includes a side wear alarm module; The side wear alarm module is used to output a side wear alarm signal for the target curve section when the evaluation result output by the side wear evaluation module indicates that the wear on the outer rail side exceeds the maintenance intervention level threshold. Furthermore, the side wear alarm module is also used to generate external rail side wear development rate information based on the accumulated side wear status evaluation results of multiple trains passing through the same target curve section. The side wear development rate information is used to assist in determining the priority and time window of rail grinding operations.

[0014] Preferably, it also includes a baseline calibration module; The baseline calibration module is used to collect feedback force time series and extract corresponding overshoot features when the target curve section is in a reference state without significant side wear, and calculate and store the baseline asymmetry ratio by collecting feedback force time series when multiple trains pass through, so that the feature extraction module can perform baseline compensation in subsequent evaluation. The acquisition module, transient detection module, feature extraction module, side wear assessment module, side wear alarm module, and baseline calibration module are integrated into the control unit of the wheel-rail force active feedback vibration absorption system. This enables the vibration absorption system to actively suppress wheel-rail force fluctuations while using its own control output parameters to achieve online detection and assessment of track defects.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention utilizes the asymmetry of the transient response of the wheel-rail force active feedback vibration absorption system when entering and exiting curves to achieve online dynamic assessment of the wear state of the outer rail side. The system directly extracts the time series of the feedback force output from the vibration absorption system's control channel on the outer rail side, extracting two transient response waveforms generated when the train enters a circular curve from a transition curve and vice versa. By calculating the asymmetry ratio of the overshoot characteristics, the wheel-rail contact point migration disturbance information contained in the vibration absorption system's control output is transformed into a quantitative criterion for the side wear state. The entire process requires no additional independent rail profile measurement sensors or dedicated detection devices, allowing track defect detection and active vibration suppression to share the same hardware platform. This solves the technical problem in existing technologies where high-frequency online dynamic assessment of outer rail side wear cannot be performed using the operating data of the train's own control system without the addition of dedicated sensors.

[0016] 2. This invention further improves the accuracy of side wear assessment through baseline compensation and a comprehensive asymmetry index. Even under new rail or no-side-wear baseline conditions, due to inherent factors such as curve superelevation settings and wheel load transfer, the transient response of the vibration-absorbing system may exhibit weak, non-pathological asymmetry when entering and exiting curves. This system compensates for the measured asymmetry ratio using a pre-calibrated baseline asymmetry ratio, eliminating the aforementioned background influence and ensuring that the compensated asymmetry index purely reflects the difference contributed by the outer rail profile asymmetry. Simultaneously, by weighted fusion of the ratios of multi-dimensional overshoot characteristics such as overshoot peak value, overshoot decay time constant, and overshoot waveform energy asymmetry, a comprehensive asymmetry index is generated. This overcomes the assessment instability that may occur with single characteristics under complex working conditions, enhancing the sensitivity of early weak side wear identification and the consistency of results.

[0017] 3. This invention also provides decision support for the transformation of rail maintenance from periodic repair to condition-based and predictive maintenance through the accumulation of data and the generation of side wear development rates. The side wear alarm module immediately outputs an alarm when the real-time assessment results exceed the maintenance intervention level. It also calculates the development rate of outer rail side wear based on accumulated side wear condition assessment data from multiple train trips and at multiple times within the same curve section. This allows track maintenance departments to scientifically determine the priority and time window for grinding operations based on the actual trend of side wear development in each curve section, rather than relying on fixed maintenance cycles. This helps optimize maintenance resource allocation, extend rail service life, and reduce life-cycle maintenance costs while ensuring safety. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall system framework of the present invention; Figure 2 This is a schematic diagram of the data acquisition and transient response detection framework of the present invention; Figure 3 This is a schematic diagram of the framework for overshoot feature extraction and asymmetric index calculation in this invention; Figure 4 This is a schematic diagram of the framework for the external rail side wear assessment and maintenance decision support of the present invention; Figure 5 This is a schematic diagram of the baseline calibration and background compensation framework of the present invention. Detailed Implementation

[0019] To facilitate understanding of the technical solution of the present invention by those skilled in the art, the technical solution of the present invention will now be further described in conjunction with the accompanying drawings.

[0020] Example 1, as Figures 1-5 As shown, the present invention provides an active feedback vibration absorption system for wheel-rail force in track corrugation sections, comprising: an acquisition module, a transient detection module, a feature extraction module, and a side wear assessment module.

[0021] The embodiments of this invention can be applied to intelligent monitoring scenarios of rail transit lines, particularly in trains or track sections equipped with wheel-rail force active feedback vibration absorption systems. While actively suppressing wheel-rail force fluctuations and mitigating corrugation, this system can also dynamically assess the wear on the outer rail side using its control output signal.

[0022] In embodiments of the present invention, the aforementioned modules work collaboratively. The acquisition module captures a high-precision feedback force time series in real time as the train passes through the target curve section. The transient detection module accurately identifies and extracts key waveform segments characterizing abrupt changes in the wheel-rail contact state from this time series. The feature extraction module quantifies the dynamic characteristics of these waveform segments and calculates their asymmetry. Finally, the side wear assessment module maps this asymmetry index to a specific external rail side wear state assessment result. The entire process requires no additional dedicated sensors and utilizes the output signal of the active control system.

[0023] Taking a specific application scenario as an example, a train equipped with this system enters a target curve section with a radius of 800 meters. The outer rail of this section has experienced some degree of side wear due to long-term service. During train operation, the acquisition module records the feedback force time series output by the outer rail side control channel in real time. At the moment the train enters the circular curve from the transition curve, due to the wear slope on the inner side of the outer rail, the wheel-rail contact point slides inward, causing a sharp change in the feedback force. The transient detection module captures and extracts the first transient response waveform corresponding to this mechanical disturbance. Similarly, at the moment the train exits the circular curve and enters the next transition curve, the contact point rises outward, and the module extracts the second transient response waveform. Subsequently, the feature extraction module extracts features such as the overshoot peak of these two waveforms and calculates their ratio. Finally, the side wear assessment module outputs the outer rail side wear amount of this section based on this ratio and a pre-calibrated mapping relationship, and determines whether maintenance intervention is required.

[0024] In one embodiment, the acquisition module includes a mileage acquisition unit, a line data unit, and a trigger recording unit, for the acquisition module and the transient detection module.

[0025] The mileage acquisition unit is used to acquire the real-time mileage information of the train; The line data unit is used to store the start and end mileage data of the curve section of the target line; The trigger recording unit is used to record the time sequence of the feedback force of the outer track side control channel when the real-time mileage information indicates that the train is about to enter the target curve section.

[0026] This approach ensures that data acquisition is event-driven rather than blind, avoiding waste of storage and computing resources, and initiating high-resolution recording only in specific segments that require evaluation.

[0027] The transient detection module includes a gradient calculation unit, a step recognition unit, and a waveform extraction unit. It detects the feedback force time series and extracts the first transient response waveform generated when the train transitions from a transition curve to a circular curve, and the second transient response waveform generated when the train transitions from a circular curve to a transition curve. The transient detection module distinguishes between the first and second transient response waveforms in the following way: By combining the real-time mileage information and line data from the acquisition module, when a step response is first detected in a predetermined curve section, it is confirmed as the first transient response waveform (entry event). Subsequently, the next step response detected when the train passes through the exit end of the same curve is confirmed as the second transient response waveform (departure event). The gradient calculation unit is used to calculate the time gradient value of the feedback force time series; The step recognition unit is used to determine the current moment as the start moment of a transient event when the time gradient value exceeds the preset step recognition condition in multiple consecutive sampling points. The condition can be a set of preset gradient thresholds to distinguish between normal gradual change of control force and force step caused by sudden change of contact point. The waveform truncation unit is used to extract the corresponding transient response waveform by extending a preset time length forward and backward from the starting time. The corresponding transient response waveform is the one that matches the detected transient event type between the first and second transient response waveforms. A gradient-based method is used to analyze the feedback force signal point by point, enabling high-time-resolution capture of the accurate moment when the wheel-rail contact state changes abruptly. This ensures the completeness of the transient waveform truncation and the accuracy of the starting point, laying the foundation for subsequent precise calculation of overshoot characteristics.

[0028] In one embodiment, the transient detection module further includes a validity verification unit; The validity verification unit acquires bogie angle change information when the train passes through a transition curve and compares the start time of the transient event with the transition curve passage period indicated by the bogie angle change information. This confirms that the start time falls within the transition curve passage period, thus verifying that the captured transient response waveform corresponds to the mechanical disturbance stage of the train entering or leaving the circular curve. By introducing independent bogie kinematic information for secondary verification of the gradient-detected transient event, false step signals caused by factors such as switch impacts and local track irregularities can be effectively filtered out. This ensures that the captured waveform is indeed triggered by the wheel-rail contact point migration event when entering or leaving the circular curve, significantly improving the physical correlation of feature extraction and the accuracy of the evaluation results.

[0029] In one embodiment, the physical meaning of the first transient response waveform and the second transient response waveform mentioned above is as follows: The first transient response waveform corresponds to the response in the feedback force control output to the step disturbance caused by the inward slippage of the wheel-rail contact point due to the wear slope on the inner side of the outer rail head when the train enters a circular curve from a transition curve. Because of the wear slope on the inner side of the outer rail head, the inward slippage of the contact point will cause an instantaneous change in the equivalent taper or rolling circle radius, which in turn triggers a step in the lateral force or eccentric moment. The active control system must respond quickly to generate a force in the opposite direction to suppress this disturbance, thus forming the first transient response waveform in the feedback force time series.

[0030] The second transient response waveform corresponds to the response in the feedback force control output of the step disturbance generated by the outward rebound of the wheel-rail contact point when the train enters a transition curve from a circular curve. This is an event where the physical process is symmetrical, but the mechanical performance may be asymmetrical due to wear. Clarifying the physical origin of these two types of waveforms and closely linking the abstract time-domain waveforms with the actual wheel-rail contact mechanics process gives the asymmetric analysis based on waveform characteristics a clear physical meaning, ensuring the rigor of subsequent evaluation logic.

[0031] In one embodiment, the overshoot feature extracted by the feature extraction module includes the overshoot peak value, the overshoot decay time constant, and the overshoot waveform energy asymmetry.

[0032] The overshoot peak value is the maximum amplitude in the transient response waveform that deviates from the steady-state value. The overshoot decay time constant is the time constant obtained by envelope fitting of the decay segment of the transient response waveform; The energy asymmetry of the overshoot waveform is the ratio between the positive energy exceeding the steady-state value and the negative energy below the steady-state value in the transient response waveform; This multi-dimensional definition can comprehensively characterize the transient response from multiple aspects, including amplitude, dynamic response speed, and energy distribution, thus capturing more precisely the mechanical differences caused by the asymmetry of the outer track profile. For example, the peak ratio alone may be affected by instantaneous noise, while analysis combining energy angle and time constant can more robustly reflect changes in system damping and stiffness characteristics caused by wear.

[0033] In one embodiment, the feature extraction module includes a ratio calculation unit and a fusion calculation unit.

[0034] The ratio calculation unit is used to calculate the ratio of the first transient response waveform and the second transient response waveform under the same type of overshoot characteristic; The fusion calculation unit is used to weight and fuse the ratios corresponding to various overshoot features when the extracted overshoot features include at least two types to generate a comprehensive asymmetry index. This comprehensive asymmetry index is input as an asymmetry ratio to the side wear evaluation module. By employing a weighted fusion strategy, the system can integrate asymmetric information from multiple dimensions to generate a more robust and representative single quantitative indicator. This avoids evaluation fluctuations caused by the instability of a single feature under certain operating conditions and enhances the system's robustness under different operating conditions and disturbances.

[0035] Specifically, the ratio calculation unit calculates the ratio of the first transient response waveform and the second transient response waveform under the same type of overshoot feature in the following manner: When the overshoot characteristic is the overshoot peak value, let the overshoot peak value of the first transient response waveform be... The overshoot peak value of the second transient response waveform is Then the overshoot peak ratio satisfy: ; When the overshoot characteristic is the overshoot decay time constant, let the overshoot decay time constant of the first transient response waveform be... The overshoot decay time constant of the second transient response waveform is Then the decay time constant is greater than satisfy: ; When the overshoot characteristic is the energy asymmetry of the overshoot waveform, let the energy asymmetry of the first transient response waveform be... The energy asymmetry of the second transient response waveform is The energy asymmetry is compared to satisfy: ; Accordingly, the fusion computing unit generates the comprehensive asymmetry index in the following manner: When overshoot features include When planting, Let the first The ratio corresponding to the overshoot feature is The corresponding weighting coefficient is Then the comprehensive asymmetry index satisfy: ; in, To The normalized mapping function performed.

[0036] The normalization function can be designed as follows: when hour This indicates that the ingress and egress responses are completely symmetrical, with no signs of wear. when When it is greater than 1, Monotonically increasing, it exacerbates the asymmetry in representation. This mathematical expression ensures that the ratios of all features are mapped to a common, physically meaningful dimension, enabling the weight allocation to truly reflect the relative importance of different features and providing a foundation for the digitization and refinement of the final evaluation results.

[0037] Weighting coefficient This is determined through a supervised calibration process. Specifically, using a calibration data set, the ratios of each feature corresponding to each calibration sample segment are determined. The result obtained after normalization mapping As input features, the known outer rail side wear of this section is used. As the target output, construct a regression model. Weight coefficients. The optimization objective is to make the comprehensive asymmetry index equal to the minimum value across all calibrated sample segments. Compared with actual wear The correlation between them is the highest, or makes it possible to achieve the desired result through subsequent mapping models. The root mean square error (RMSE) between the predicted wear amount and the actual wear amount is minimized. Taking the minimization of the RMSE as an example, the optimization objective function is: ; in To determine the total number of sample segments, the optimal weighting coefficients can be found using gradient descent or grid search methods. The optimization results are then embedded into the fusion computing unit.

[0038] In one embodiment, the feature extraction module further includes a baseline compensation unit; The baseline compensation unit is used to obtain the baseline asymmetry ratio of the same target curve section under the standard track profile when multiple trains pass through, and to use the baseline asymmetry ratio to compensate the current calculation result when calculating the asymmetry ratio, so as to eliminate the background influence caused by curve superelevation and wheel load transfer.

[0039] This means that even if the track profile is perfectly symmetrical, the response when entering and exiting a curve may still exhibit inherent, non-pathological, slight asymmetry due to inherent mechanical distribution effects such as curve superelevation settings and wheel load transfer caused by vehicle passage. Baseline compensation can eliminate this background effect. Specifically, let the ratio of uncompensated asymmetry output by the feature extraction module be... The baseline asymmetry ratio is Then the asymmetry ratio after compensation satisfy: The side wear assessment module will The asymmetry ratio is used for dynamic evaluation. The implementation of this mechanism ensures that the final asymmetry ratio used for evaluation... It purely reflects the difference contributed by the asymmetry of the outer rail profile (i.e., side wear), thereby elevating the features extracted by this invention from a qualitative correlation to a quantitatively reliable level, significantly improving the accuracy of side wear assessment and the ability to identify early and subtle side wear.

[0040] In one embodiment, the side wear assessment module includes a mapping model storage unit and an assessment output unit; The mapping model storage unit is used to store pre-built mapping models. The mapping model defines the correspondence between the asymmetry ratio and the wear amount on the outer rail side. The mapping model is also classified and built according to the curve radius, rail material and total weight.

[0041] The classification of these factors ensures the universality and specificity of the mapping relationship. The evaluation output unit is used to call the corresponding mapping model based on the actual track parameters of the target curve segment, and inputs the asymmetry ratio into the called mapping model to output the outer rail side wear of the target curve segment. It also compares the outer rail side wear with a preset maintenance intervention level threshold, outputting the evaluation result of the outer rail side wear state of the target curve segment. Here, the compensated asymmetry ratio is set to... The mapping model called is The wear on the outer rail side satisfy: By calling the mapping model established by classification, the system can take into account the impact of different curve types, rail hardness and wear resistance, and cumulative transport volume on the rate and characteristics of side wear development. It transforms abstract indices into intuitive millimeter-level wear amounts in engineering and directly compares them with maintenance standards, achieving a seamless connection from data analysis to decision support.

[0042] To further illustrate the mapping model In this embodiment, the mapping model is established through a combination of offline calibration and online invocation.

[0043] First, on the target line or a test section with similar conditions, several curves with different degrees of outer rail side wear were selected as calibration sample sections. Using offline precision measurement methods such as a rail profile measuring instrument, the accurate outer rail side wear of each calibration sample section was obtained and denoted as... The system involves arranging for trains equipped with it to pass through each calibration sample section multiple times. Following the method described in this invention, the compensated asymmetry ratio corresponding to each calibration sample section is obtained for each passage. The average of multiple passes through the same section is used to form a set of calibration data. Based on the data set of all calibrated sample segments, a least squares method was used for multinomial regression fitting. For specific classification conditions, namely, calibration data with the same curve radius, the same rail material, and similar total passing weight, a mapping function was established. .

[0044] In one embodiment, the system further includes a side wear alarm module; The side wear alarm module is used to output a side wear alarm signal for the target curve section when the evaluation result output by the side wear assessment module indicates that the side wear of the outer rail exceeds the maintenance intervention level threshold.

[0045] Furthermore, the side wear alarm module is also used to generate information on the development rate of outer rail side wear based on the accumulated side wear status assessment results of multiple trains passing through the same target curve section. This side wear development rate information is used to help determine the priority and time window of rail grinding operations.

[0046] By introducing alarm and development rate analysis functions, the system not only provides real-time status snapshots but also performs trend predictions. This enables maintenance departments to shift from "planned maintenance" to "condition-based maintenance" and even "predictive maintenance," scientifically prioritizing grinding operations based on the wear development rate and current status of different sections, optimizing resource allocation, effectively extending the service life of rails, and ensuring traffic safety.

[0047] In one embodiment, the system further includes a baseline calibration module; The baseline calibration module is used to collect feedback force time series during multiple train passes and extract corresponding overshoot features under the condition that the target curve section is in a reference state without significant side wear. It calculates and stores the baseline asymmetry ratio so that the feature extraction module can perform baseline compensation in subsequent evaluations.

[0048] The acquisition module, transient detection module, feature extraction module, side wear assessment module, side wear alarm module, and baseline calibration module are integrated into the control unit of the wheel-rail force active feedback vibration absorption system. This highly integrated design enables the vibration absorption system to actively suppress wheel-rail force fluctuations while simultaneously using its own control output parameters to achieve online detection and assessment of track defects. This multifunctional approach significantly reduces the hardware cost and complexity of system deployment.

[0049] As a specific implementation method, baseline calibration can be performed first on newly polished rails or sections known to be in good condition to establish the purest dynamic baseline reference. Through integration with the system control unit, the algorithms of these modules can run on a digital signal processor or microcontroller.

[0050] It should be noted that in some optional implementations, the step identification condition can be a dynamic threshold, which is adaptively adjusted based on the statistical characteristics (such as the root mean square value) of the preceding steady-state value of the feedback force time series to cope with changes in the control force baseline noise at different operating speeds. The truncation time length of the transient response waveform can be dynamically adjusted according to the train's operating speed to ensure that the truncation waveform completely encompasses the entire transient process without introducing excessive steady-state signals. For the envelope fitting of the overshoot decay time constant, a least-squares fitting of a first-order exponential function can be performed after obtaining the envelope using the Hilbert transform. For the normalized mapping function... A simple implementation is In other alternative implementations, to more comprehensively describe the signal characteristics, the types of overshoot features may also include the number of oscillations in the overshoot process or the skewness characteristics of the transient response waveform.

[0051] In another implementation, the normalized mapping function A nonlinear mapping form can be used. Let the first... overshoot eigenvalue The typical range of variation is This range is predetermined through statistical analysis of a large amount of historical operational data. The normalized mapping function is defined as: ; in A coefficient used to control the steepness of the mapping curve. This function satisfies the condition that... hour ,when hour It increases monotonically and gradually approaches saturation. Parameter The selection of is such that the mapped The objective is to ensure that the values ​​have a similar numerical distribution range across all features, which is determined through statistical analysis of the calibration dataset. The advantage of this nonlinear mapping method is that when the ratio... The mapping function exhibits high sensitivity at small deviations around 1 (corresponding to early, weak lateral wear), while larger deviations are moderately compressed to prevent individual extreme values ​​from excessively dominating the fusion results.

[0052] It should be noted that the key technical terms involved in this invention are explained as follows: Wheel-rail force active feedback vibration absorption system: refers to a system installed on a vehicle or track that generates controllable force (i.e. feedback force) through active control strategy to offset or suppress harmful interaction forces between wheel and rail in real time.

[0053] Feedback force time series: refers to a series of control force commands or measured values ​​continuously output in chronological order by the control channel corresponding to the outer rail side of the vibration absorption system. This time series is a direct reflection of the system's dynamic adjustment process.

[0054] Transient response waveform: refers to the time-amplitude waveform segment that appears when the train's running state changes drastically (such as from a straight line to a transition curve, or from a transition curve to a circular curve), and the feedback force transitions from one steady state to another (or an impact oscillation occurs) in a short period of time.

[0055] Overshoot characteristics: These are parameterized indicators used to quantify the degree to which a transient response waveform deviates from its final steady-state value and its dynamic characteristics, such as the magnitude of the peak value exceeding the steady-state value, the rate of decay, and the energy distribution.

[0056] Asymmetry ratio: This refers to a scalar value obtained by calculating the ratio of the overshoot characteristics generated when a train enters and leaves a circular curve. The further this ratio deviates from 1, the greater the difference in entry and exit responses caused by the asymmetry of the outer track profile.

[0057] The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.

Claims

1. A wheel-rail force active feedback vibration absorption system for track corrugation sections, characterized in that, include: The acquisition module is used to acquire the time series of feedback force output by the wheel-rail force active feedback vibration absorption system corresponding to the outer rail side control channel in real time when the train passes through the target curve section; The transient detection module is used to detect the feedback force time series and extract the first transient response waveform generated when the train enters the circular curve from the transition curve, and the second transient response waveform generated when the train enters the transition curve from the circular curve. The feature extraction module is used to extract the overshoot features of the first transient response waveform and the second transient response waveform respectively, and to calculate the ratio of the extracted overshoot features to obtain the asymmetry ratio. The asymmetry ratio is used to characterize the degree of difference in the transient response of the feedback force when entering the curve and leaving the curve due to the asymmetry of the outer rail head profile. The side wear assessment module is used to dynamically assess the side wear status of the outer rail in the target curve section based on the pre-established mapping relationship between the asymmetry ratio and the wear status of the outer rail side.

2. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 1, characterized in that, The acquisition module includes a mileage acquisition unit, a route data unit, and a trigger recording unit; The mileage acquisition unit is used to acquire the real-time mileage information of the train, the line data unit is used to store the start and end mileage data of the curve section of the target line, and the trigger recording unit is used to trigger the recording of the feedback force time sequence of the outer track side control channel when the real-time mileage information indicates that the train is about to enter the target curve section. The transient detection module includes a gradient calculation unit, a step recognition unit, and a waveform truncation unit; The gradient calculation unit is used to calculate the time gradient value of the feedback force time series. The step identification unit is used to determine that the current time is the start time of the transient event when the time gradient value exceeds the preset step identification condition within multiple consecutive sampling points. The waveform truncation unit is used to extract the corresponding transient response waveform by extending a preset time length forward and backward based on the start time. The corresponding transient response waveform is the one that matches the detected transient event type between the first transient response waveform and the second transient response waveform.

3. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 2, characterized in that, The transient detection module also includes a validity verification unit; The validity verification unit is used to obtain bogie angle change information when the train passes through the transition curve, and compare the start time of the transient event with the transition curve passing time indicated by the bogie angle change information to confirm that the start time is within the transition curve passing time, so as to confirm that the captured transient response waveform corresponds to the mechanical disturbance stage of the train entering the circular curve and leaving the circular curve.

4. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 1, characterized in that, The first transient response waveform corresponds to the response in the feedback force control output of the step disturbance caused by the wear slope on the inner side of the outer rail head causing the wheel-rail contact point to slip inward when the train enters the circular curve from the transition curve; The second transient response waveform corresponds to the response of the step disturbance generated by the outward rise of the wheel-rail contact point when the train enters the transition curve from the circular curve.

5. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 1, characterized in that, The overshoot characteristics are the overshoot peak value, the overshoot decay time constant, and the overshoot waveform energy asymmetry. Wherein, the overshoot peak value is the maximum amplitude in the transient response waveform that deviates from the steady-state value; The overshoot decay time constant is a time constant obtained by envelope fitting of the decay segment of the transient response waveform; The overshoot waveform energy asymmetry is the ratio between the positive energy exceeding the steady-state value and the negative energy below the steady-state value in the transient response waveform.

6. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 5, characterized in that, The feature extraction module includes a ratio calculation unit and a fusion calculation unit; The ratio calculation unit is used to calculate the ratio of the first transient response waveform and the second transient response waveform under the same type of overshoot feature; The fusion calculation unit is used to weight and fuse the ratios corresponding to various overshoot features when the extracted overshoot features include at least two types to generate a comprehensive asymmetry index. The comprehensive asymmetry index is then input as an asymmetry ratio to the side wear assessment module. The ratio calculation unit calculates the ratio of the first transient response waveform and the second transient response waveform under the same type of overshoot feature in the following manner: When the overshoot characteristic is the overshoot peak value, let the overshoot peak value of the first transient response waveform be... The overshoot peak value of the second transient response waveform is Then the overshoot peak ratio satisfy: ; When the overshoot characteristic is the overshoot decay time constant, let the overshoot decay time constant of the first transient response waveform be... The overshoot decay time constant of the second transient response waveform is The decay time constant is compared to satisfy: ; When the overshoot characteristic is the energy asymmetry of the overshoot waveform, let the energy asymmetry of the first transient response waveform be... The energy asymmetry of the second transient response waveform is The energy asymmetry is compared to satisfy: ; The fusion computing unit generates the comprehensive asymmetry index in the following manner: When overshoot features include When planting, Let the first The ratio corresponding to the overshoot feature is The corresponding weighting coefficient is Then the comprehensive asymmetry index satisfy: ; in, To The normalized mapping function performed.

7. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 6, characterized in that, The feature extraction module also includes a baseline compensation unit; The baseline compensation unit is used to obtain the baseline asymmetry ratio of the same target curve section under the standard track profile when multiple trains pass through, and to use the baseline asymmetry ratio to compensate the current calculation result when calculating the asymmetry ratio, so as to eliminate the background influence caused by curve superelevation and wheel load transfer. Wherein, the uncompensated asymmetry ratio output by the feature extraction module is denoted as The baseline asymmetry ratio is Then the asymmetry ratio after compensation satisfy: ; The side wear assessment module will It is used as an asymmetric ratio for dynamic evaluation.

8. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 1, characterized in that, The side wear assessment module includes a mapping model storage unit and an assessment output unit; The mapping model storage unit is used to store pre-built mapping models. The mapping model defines the correspondence between the asymmetry ratio and the wear amount on the outer rail side. The mapping model is constructed according to the curve radius, rail material and total weight. The evaluation output unit is used to call the corresponding mapping model according to the actual line parameters of the target curve section, and input the asymmetry ratio into the called mapping model to output the outer rail side wear of the target curve section, and compare the outer rail side wear with the preset maintenance intervention level threshold to output the evaluation result of the outer rail side wear state of the target curve section. Wherein, let the asymmetry ratio after compensation be ? The mapping model called is The wear amount on the outer rail side satisfy: .

9. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 1, characterized in that, It also includes a side wear alarm module; The side wear alarm module is used to output a side wear alarm signal for the target curve section when the evaluation result output by the side wear evaluation module indicates that the wear on the outer rail side exceeds the maintenance intervention level threshold. Furthermore, the side wear alarm module is also used to generate external rail side wear development rate information based on the accumulated side wear status evaluation results of multiple trains passing through the same target curve section. The side wear development rate information is used to assist in determining the priority and time window of rail grinding operations.

10. The wheel-rail force active feedback vibration absorption system for track corrugation sections according to claim 1, characterized in that, It also includes a baseline calibration module; The baseline calibration module is used to collect feedback force time series and extract corresponding overshoot features when the target curve section is in a reference state without significant side wear, and calculate and store the baseline asymmetry ratio by collecting feedback force time series when multiple trains pass through, so that the feature extraction module can perform baseline compensation in subsequent evaluation. The acquisition module, transient detection module, feature extraction module, side wear assessment module, side wear alarm module, and baseline calibration module are integrated into the control unit of the wheel-rail force active feedback vibration absorption system. This enables the vibration absorption system to actively suppress wheel-rail force fluctuations while using its own control output parameters to achieve online detection and assessment of track defects.

Images