Railway wagon brake shoe displacement and damage characteristic oriented structure defect identification method

Abstract

The invention relates to the technical field of damage identification, in particular to a structural defect identification method for brake shoe displacement and damage characteristics of a railway wagon, which comprises the following steps of: in the whole life cycle of the railway wagon, dividing a braking process into a plurality of stages, and performing hierarchical filing according to working condition types and service stages, constructing a brake shoe multi-working-condition structure response file; through longitudinal comparison along a time axis in the same working condition layer and transverse comparison among different working condition layers, screening out a brake shoe ectopic evolution mark and an initial brake shoe damage evolution mark which are sensitive to a structure state; encoding all evolution marks according to a spatial position, a working condition layer and a service stage to form a brake shoe ectopic and damage structure fingerprint sequence; and outputting an identification result including the structure defect type and the severity level, and generating a maintenance priority. According to the method, a space-time coupling mechanism of acoustic emission energy and temperature gradient is introduced to recognize early microcrack propagation, and the recognition sensitivity of recessive fatigue and complex damage modes is improved.

Description

Technical Field

[0001] This invention relates to the field of damage identification technology, and in particular to a method for identifying structural defects such as misalignment and damage characteristics of brake shoes for railway freight cars. Background Technology

[0002] As heavy-load, high-speed transportation equipment, the braking system of railway freight cars is a key link in ensuring operational safety. As the direct friction element in the wheel braking process, the brake shoe's structural condition has a direct impact on braking performance, braking balance, and wheelset wear. During long-term service, the brake shoe is prone to displacement, uneven wear, material fatigue, and even structural damage due to repeated high-energy friction, load fluctuations, and installation deviations.

[0003] Currently, the inspection of railway freight car brake shoes largely relies on regular manual maintenance or visual surface wear checks. Traditional inspection methods primarily focus on geometric parameters such as brake shoe thickness and wheel-rail clearance, making it difficult to capture the changes in the mechanical, thermal, and vibration response characteristics of the structure under different operating conditions, and failing to provide early warnings of internal structural deterioration or changes in installation status. Furthermore, the operating environment of railway freight cars is complex and variable, with significant differences between the load conditions on the brake shoes and the track conditions. Existing methods lack a comprehensive life-cycle record that correlates structural response with operating conditions, resulting in fragmented data that is difficult to use for structural evolution analysis. Minor installation deviations, loosening, or internal cracks are difficult to detect in the early stages using traditional methods, and once abrupt changes occur, irreversible damage often results. Current technologies lack the ability to sensitively capture and model the evolutionary trends of these low-level, latent anomalies. Summary of the Invention

[0004] This invention provides a structural defect identification method for the dislocation and damage characteristics of brake shoes in railway freight cars. It can continuously collect multi-dimensional response data throughout the entire life cycle, extract key indicators reflecting the evolution of brake shoe dislocation and damage, establish a standardized structural fingerprint sequence, and combine it with a defect pattern library to determine the type and output intelligent maintenance strategies, thereby realizing the transformation from periodic manual inspection to state perception + intelligent decision-making.

[0005] A method for identifying structural defects related to misalignment and damage characteristics of railway freight car brake shoes includes the following steps: S1. Construct a multi-condition structural response archive for brake shoes: During the entire life cycle of railway freight cars, based on the braking process under multiple typical working conditions, collect the vibration response, deformation response and temperature response of the brake shoes, divide the braking process into multiple stages and archive them in layers according to working condition type and service stage to construct a multi-condition structural response archive for brake shoes. S2, Establish brake shoe displacement and damage structure fingerprints: Using the brake shoe multi-condition structural response file as input, longitudinal comparison along the time axis within the same condition layer and lateral comparison between different condition layers are performed to filter out brake shoe displacement evolution markers and initial brake shoe damage evolution markers that are sensitive to structural state: Among them, for the selected brake shoe damage evolution markers, a coupling analysis of acoustic emission signal energy release rate and brake shoe surface temperature gradient is introduced to identify the microcrack propagation behavior inside the brake shoe material due to long-term alternating load accumulation, and obtain a comprehensive brake shoe damage evolution marker characterizing the fatigue state of the material; all evolution markers are encoded according to spatial location, working condition level and service stage to form a fingerprint sequence of brake shoe displacement and damage structure. S3, Identify structural defects and output maintenance priorities: Match the fingerprint sequence of the misplaced brake shoe and damaged structure with a pre-established structural defect pattern library, output identification results including structural defect type and severity level, and generate maintenance priorities; when a persistent low-level structural defect trend is detected, the trend is used to adaptively update the health boundary to improve the sensitivity of early defect identification.

[0006] Optionally, throughout the entire life cycle of the railway freight car, vibration response is collected by arranging an acceleration sensor at the brake shoe mounting base, deformation response is collected by arranging strain gauges at key stress locations on the brake beam, and temperature response is collected by arranging an infrared temperature sensor at the brake shoe back plate. The brake cylinder pressure signal is also collected simultaneously as a benchmark for dividing the working conditions.

[0007] Optionally, based on the characteristics of the brake cylinder pressure signal, the braking process is divided into a brake establishment stage, a pressure holding stage, and a release stage, and the pressure holding stage is selected as the brake shoe action segment.

[0008] Optionally, the vibration response, deformation response, and temperature response data are classified into first-level categories according to empty / loaded vehicle operating conditions and straight road / long slope operating conditions. Within each operating condition level, the service stages are further divided according to vehicle mileage, including the initial break-in stage, stable service stage, and end-of-wear stage. Time-domain and frequency-domain features are extracted from the response data of multiple brake shoe action segments under the same operating condition and the same service stage, including the root mean square value of vibration acceleration, peak strain, and temperature rise slope, to construct a multi-operating-condition structural response profile of brake shoes with unified feature dimensions.

[0009] Optionally, the longitudinal comparison along the time axis within the same working condition layer includes calculating the growth rate of the root mean square value of vibration acceleration, the shift trend of the strain peak value, and the cumulative change of the temperature rise slope, and marking the abnormal change points that exceed the normal wear range as brake shoe displacement evolution markers.

[0010] Optionally, the lateral comparison between different working conditions includes identifying the load sensitivity coefficient by the difference in strain peak value under loaded and empty vehicle working conditions, identifying the braking intensity sensitivity coefficient by the difference in temperature rise slope under long slope and straight road working conditions, and marking the change point of the sensitivity coefficient exceeding the threshold as the brake shoe damage evolution marker.

[0011] Optionally, for the selected initial brake shoe damage evolution markers, acoustic emission signals collected by acoustic emission sensors arranged on the back plate of the brake shoe are introduced, the energy release rate of the acoustic emission signals is calculated, and the temperature gradient distribution on the surface of the brake shoe collected by infrared temperature sensors is combined to establish a coupled analysis model of energy release rate and temperature gradient. When the abrupt change point of energy release rate coincides with the abnormal temperature gradient region in time and space, it is determined to be the microcrack propagation behavior inside the brake shoe material, thus obtaining a comprehensive brake shoe damage evolution marker characterizing the fatigue state of the material.

[0012] Optionally, S2 further includes encoding the brake shoe displacement evolution marker and the comprehensive brake shoe damage evolution marker with three-dimensional codes according to the spatial installation location, working condition level code and service stage number of the brake shoe, to form a brake shoe displacement and damage structure fingerprint sequence with spatiotemporal characteristics.

[0013] Optionally, the structural defect pattern library is constructed based on test bench data and typical faulty prototype vehicle data, including typical feature patterns of brake beam bending, tie rod loosening, suspension rod deformation, and bracket cracking; by calculating the similarity between the structural fingerprint sequence and each defect pattern, the type of structural defect is identified, and the severity level of the structural defect is determined according to the deviation of the abnormal feature amplitude from the standard threshold. Maintenance priorities are generated based on the severity level and development trend of structural defects. Defects with high severity level and rapid deterioration are marked as urgent maintenance, defects with medium severity level but stable are marked as planned maintenance, and low-level defects that persist are marked as observation maintenance.

[0014] Optionally, S3 further includes: when the same brake shoe location shows a low-level structural defect trend in three consecutive maintenance cycles, extracting the characteristic parameters of the trend, including vibration growth rate, temperature change slope and acoustic emission energy accumulation, for adaptively updating the health boundary threshold, and writing the updated health boundary into the brake shoe multi-condition structural response file.

[0015] The beneficial effects of this invention are: This invention constructs a multi-condition structural response profile covering empty / loaded cars, slopes / level tracks, and different service stages by deploying various sensors such as vibration, strain, temperature, and acoustic emission throughout the entire life cycle of railway freight cars. Based on longitudinal trend analysis under the same working condition and lateral sensitivity comparison across working conditions, it extracts brake shoe displacement evolution markers and comprehensive damage evolution markers. This not only captures minute structural changes such as positional shifts and abnormal load responses, but also introduces a spatiotemporal coupling mechanism of acoustic emission energy and temperature gradient to identify early microcrack propagation, thereby improving the sensitivity of identifying latent fatigue and complex damage modes.

[0016] This invention constructs a traceable structural fingerprint sequence by three-dimensionally encoding the spatial location, working condition level, and service stage of structural evolution markers. This fingerprint sequence is then matched with a multi-dimensional feature library of structural defect patterns built based on bench tests and prototype vehicle measurements. This enables the identification of typical defects such as brake beam bending, tie rod loosening, suspension rod deformation, and bracket cracks. Furthermore, based on a deviation calculation model between response features and standard templates, defects are classified into three levels: severe, moderate, and minor. This provides a quantitative basis for subsequent maintenance decisions and offers a clear diagnosis, classification, and handling process, improving the automation and accuracy of fault determination.

[0017] This invention addresses the problem of mild but persistent defects often overlooked in traditional methods by proposing a health boundary self-updating mechanism based on multi-cycle low-level defect trend analysis. It extracts trend features including vibration growth rate, temperature slope change, and acoustic emission energy accumulation, and weights these features into the health judgment threshold to achieve dynamic correction of the response criteria. This mechanism enables the system to learn and automatically optimize the judgment criteria according to the actual operating state, thereby improving the ability to identify gradual and non-sudden structural degradation. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the identification method flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the establishment of fingerprints for misplaced and damaged brake shoes according to an embodiment of the present invention. Detailed Implementation

[0020] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. For some well-known technologies, those skilled in the art may also use other alternative methods to implement the invention. Moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0021] like Figures 1-2 As shown, a structural defect identification method for misaligned and damaged brake shoes of railway freight cars includes the following steps: S1. Construct a multi-condition structural response archive for brake shoes: During the entire life cycle of railway freight cars, based on the braking process under multiple typical working conditions, collect the vibration response, deformation response, and temperature response of the brake shoes, divide the braking process into multiple stages, and archive them in layers according to working condition type and service stage to construct a multi-condition structural response archive for brake shoes.

[0022] Specifically, it includes: Throughout the entire life cycle of railway freight cars, a multi-condition structural response profile for brake shoes is established using the following sensor configurations and data acquisition methods: The full life cycle of a railway freight car is the entire service life from the time it leaves the factory and is put into use until it is scrapped and retired. Within this cycle, the freight car goes through several stages, including: Initial break-in phase: When newly put into use, components gradually adapt to the load and operating environment; Stable service phase: The main operating phase, characterized by high usage frequency and relatively low failure rate; Wear and aging stage: As operating time and mileage accumulate, key components gradually show signs of wear, fatigue, or aging, requiring more frequent maintenance. Retirement stage: Approaching or reaching the design life, the vehicle's performance declines or costs increase, making it unsuitable for continued use.

[0023] The term "full life cycle" is used to illustrate that the structural response data of the brake shoe covers the entire life cycle, multiple operating conditions, and multiple states, from new vehicles to old vehicles, from light loads to heavy loads, and from flat roads to slopes. This provides a data foundation for the subsequent establishment of accurate defect identification models and health boundary update mechanisms.

[0024] S11, Response Signal Acquisition Configuration: A triaxial accelerometer is installed at the brake shoe mounting base to collect the vibration response of the brake shoe during braking. Strain gauges were placed at key stress locations on the brake beam to collect the deformation response of the braking force transmission path; An infrared temperature sensor is installed on the back plate of the brake shoe to collect the temperature response during the contact and friction process between the brake shoe and the wheel. Pressure sensors are synchronously deployed to collect brake cylinder pressure signals, which serve as the benchmark for dividing the braking process conditions.

[0025] S12, Braking condition segment division: based on brake cylinder pressure signal Based on the changing trend, the complete braking process can be divided into three stages: Braking establishment phase: ; Pressure maintenance phase: ; Relief phase: ; The pressure holding phase was selected as the core segment of the brake shoe action, and the response data within this phase was extracted for subsequent analysis. This represents the curve showing the change of the brake cylinder pressure signal over time. This is a threshold used to determine pressure rise and fall, with a value ranging from 0.05 to 0.1 MPa / s. Based on the actual characteristics of the air pressure rise / fall rate in truck braking systems, the typical brake air pressure change rate is approximately 0.2–0.5 MPa / s. Therefore, 10% to 20% is selected as the threshold to distinguish between changing and steady states. It is a small change tolerance threshold used to identify the pressure holding phase, with a value ≤0.01 MPa / s. Considering the small pressure fluctuations under actual operating conditions (such as sensor noise or air pressure control fine-tuning), this small value is set to identify the stable section of the pressure holding phase.

[0026] During each braking process of a truck, the change in air pressure within the brake cylinder directly reflects the braking intensity. This air pressure (i.e., the brake cylinder pressure signal) The pressure changes according to the execution of the braking command, typically exhibiting a trend of first rising, then holding, and then falling. This is achieved by applying the first derivative of the pressure signal. By performing calculations, it can be determined which stage the braking process is in: Braking establishment phase: If the derivative is significantly positive, it indicates that the pressure is rising rapidly; Pressure holding phase: If the derivative is close to zero, it indicates that the pressure is stabilizing; Relief phase: If the derivative is significantly negative, it indicates that the pressure is beginning to decrease and the braking is being released.

[0027] During the pressure holding phase, the brake shoe maintains stable contact with the wheel and bears most of the braking force; this is the phase where brake shoe wear, displacement, and damage are most significant. Compared to the instantaneous impact of the build-up phase and the unloading process of the relief phase, the signals from the pressure holding phase better reflect the true working state of the brake shoe structure under continuous load, including the stable vibration mode, the limiting strain value, and the continuous temperature rise trend. By selecting only the vibration, strain, and temperature response data within this time window of the pressure holding phase as the analysis object, and extracting effective features characterizing the brake shoe state (root mean square, peak strain, temperature rise rate, etc.), rather than data from the entire braking process, this strategy helps to remove stage-specific noise, focus on the mechanism itself, and improve identification accuracy.

[0028] S13, Data Archiving and Hierarchical Classification: Archive response data in a two-level classification system based on operating conditions and service stages. S131, working condition classification: classified by load conditions into empty car working condition and loaded car working condition; classified by track conditions into straight road working condition and long and steep slope working condition. S132, Service Phase Classification: Initial break-in period (0–200,000 km); Stable service phase (200,000–800,000 km); Late wear stage (800,000–1,000,000 kilometers and above).

[0029] The purpose of S13 is to systematically organize and hierarchically archive the large amount of brake shoe response data collected according to certain rules, thereby constructing a clear and highly comparative multi-condition structural response archive of brake shoes, providing high-quality input for subsequent identification models. Specifically: I. First-level classification by operating conditions: This is a preliminary classification of the data based on the environmental conditions in which the brake shoes operate, specifically including the following two dimensions: Load condition classification: Empty vehicle condition: The operating state of a freight truck when it is not loaded or only lightly loaded, and the force on the brake shoes is relatively small when braking; Heavy vehicle operating conditions: When a truck is fully loaded or nearly fully loaded, the braking intensity is high, which leads to more significant wear on the brake shoes.

[0030] Line condition classification: Straight-line road conditions: On conventional lines, such as station areas and plains, braking behavior is relatively mild; Long slope conditions: such as downhill in mountainous areas and long-distance gradient lines, the braking time is long, the heat accumulation effect is obvious, and the brake shoes are more prone to high temperature fatigue and damage.

[0031] The aim is to separate the data according to different operating environments, which helps to identify the changing patterns of brake shoe response under specific operating conditions and avoids analysis distortion caused by mixing different operating scenarios.

[0032] II. Within the operating condition level, a second level of classification is performed based on the service stage: Under each operating condition, the service life stage of the brake shoe is further subdivided, specifically into: Initial break-in period (0–200,000 km): When the brake shoes are first put into use, the material surface has not yet fully adhered to the wheel, and the response characteristics deviate from the stable state; slight displacement is more likely to occur during the structural rigidity adjustment process.

[0033] Stable service stage (200,000–800,000 km): The brake shoes have reached a stable working state, and the response characteristics are representative; this is the core stage for extracting standard response characteristics under normal conditions.

[0034] Late wear stage (800,000–1,000,000 km and above): The brake shoe wear is close to the limit, structural fatigue accumulates, and potential damage signs such as high-temperature local cracks and positional displacement are likely to occur.

[0035] The aim is to reveal the trend of the evolution of the brake shoe structure over time by classifying it by service stage, and to support the temporal logic and hierarchical tracking of subsequent identification of displacement and damage processes.

[0036] S14, Response Feature Extraction and Unified Characterization: The selected brake shoe response data is feature extracted and uniformly represented into a standardized vector form for easy subsequent comparison, modeling, and defect identification. In each operating condition (empty / loaded, straight / slope) and each service stage (break-in, stabilization, wear-out), the truck recorded a large amount of brake shoe action segment response data during multiple braking processes. This response data mainly comes from three types of sensors: acceleration sensors, strain gauges, and temperature sensors. For ease of analysis, three key features are extracted from the three types of response data in each action segment: Vibration intensity index: The root mean square value of the triaxial vibration acceleration signal is calculated, which represents the overall vibration energy of the brake shoe during the braking process and reflects the dynamic load response intensity of the structure during braking.

[0037] Deformation strength index: Extract the maximum value (peak value) of the strain signal, which represents the maximum mechanical deformation of the brake shoe or brake beam in this segment, and is used to determine whether there is structural rigidity degradation or abnormal stress.

[0038] Thermal excitation index: The slope of the temperature signal is calculated, reflecting the rate of temperature rise per unit time during the friction between the brake shoe and the wheel. This feature helps to identify thermal fatigue or material damage caused by high temperature.

[0039] These three features together form a feature vector, a data structure with a unified format, used to represent the structural response state of the brake shoe during a braking process. Each feature vector clearly identifies the type of operating condition, the service stage, and the segment number for which data was acquired.

[0040] Specifically, this involves extracting features from the response data of multiple brake shoe action segments under each type of operating condition and each service stage (only selecting the response data during the pressure holding phase), forming a feature vector of a unified dimension. , representing the response characteristics of the i-th operating condition, the j-th service stage, and the k-th action segment: ; in, Indicates triaxial vibration acceleration signal The root mean square value is used to measure vibration intensity. This represents the peak value of the strain signal, the maximum response output of the strain gauge, used to characterize the peak value of structural deformation. Represents temperature signal The rising slope during the pressure holding phase is used to assess the rate of frictional heat generation.

[0041] Ultimately, all Multi-dimensional archiving is performed based on operating condition type, service stage, and response characteristics to construct a multi-condition structural response archive for brake shoes. This provides standardized basic data for subsequent structural fingerprint extraction and defect identification.

[0042] S2, Establish brake shoe displacement and damage structure fingerprints: Using the brake shoe multi-condition structural response file as input, longitudinal comparison along the time axis within the same condition layer and lateral comparison between different condition layers are performed to filter out brake shoe displacement evolution markers and initial brake shoe damage evolution markers that are sensitive to structural state: Among them, for the selected brake shoe damage evolution markers, a coupling analysis of acoustic emission signal energy release rate and brake shoe surface temperature gradient is introduced to identify the microcrack propagation behavior inside the brake shoe material due to long-term alternating load accumulation, and obtain a comprehensive brake shoe damage evolution marker characterizing the fatigue state of the material; all evolution markers are encoded according to spatial location, working condition level and service stage to form a fingerprint sequence of brake shoe displacement and damage structure.

[0043] S2 uses the brake shoe multi-condition structural response file Using this as input, comparative analyses are performed within the same operating condition layer and between different operating condition layers to extract key markers characterizing the evolution of the brake shoe structure. Specifically: S21. Longitudinal comparison within the same working condition layer to extract ex-situ evolution markers: Under the same working condition type, by longitudinally analyzing the changing trends of characteristic indicators in each service stage, abnormal evolution phenomena are identified and marked as signals that may indicate structural ex-situation, for subsequent structural defect identification. During service, the structural state of railway freight car brake shoes changes with time and usage intensity. To track these changes, this scheme analyzes the evolution trends of the following three key indicators under fixed working conditions, such as heavy vehicles + slope, in chronological order, i.e., as the service stage progresses: The rate of increase of the root mean square value of vibration acceleration: measures whether the vibration experienced by the brake shoe gradually increases over time. If the vibration energy increases significantly in a certain stage compared to the previous stage, it may mean that there is an abnormal contact or positional change between the brake shoe and the wheelset.

[0044] The shift trend of strain peak value: reflects whether there is a sudden change in structural deformation. If the strain peak value changes drastically between adjacent stages, it may mean that the brake shoe has been displaced (slightly moved in position or deflected at an angle), and the influence transmission path.

[0045] The cumulative change in the slope of temperature rise: track whether the trend of frictional heat generation of the brake shoe is abnormally aggravated. If the cumulative temperature rise rate is much higher than the change value during the normal wear period, it indicates the existence of a continuous bias or a local high friction zone.

[0046] These three indicators reflect the rate of change of the structure in three dimensions: dynamics (vibration), mechanics (deformation), and thermal (temperature), respectively. When these changes exceed the preset normal wear range, the response is considered to be structurally abnormal and no longer belongs to the normal aging process, but rather to abnormal evolution that may be caused by displacement or other hidden problems.

[0047] Once any of the following conditions are met in a certain stage of the data: The rate of change of vibration acceleration is too large. The strain peak shifted drastically. The rate of cumulative temperature rise changed abruptly. The data at this stage is then labeled as a gate shoe displacement evolution marker to support the subsequent inclusion of this anomaly in the structural fingerprint, for identification and backtracking by intelligent diagnostic algorithms.

[0048] The specific scheme is as follows: Under a fixed operating condition type i, the feature vector sequence is processed along the time axis (increasing with the service stage number j). Evolutionary trend analysis of key indicators in the data: Vibration acceleration growth rate: ; indicates the rate of increase in vibration intensity per unit mileage of the brake shoe between the j-th and j+1-th service stages, used to determine whether the vibration is abnormally increasing and to reflect changes in structural contact.

[0049] Strain peak offset trend: This indicates the range of change in the maximum strain value of the brake shoe between adjacent service stages. It is used to identify whether there is an abrupt change in the structural stress path, which may indicate brake shoe displacement or local loosening.

[0050] The cumulative change in the slope of temperature rise: ; represents the sum of the temperature rise rates up to stage j, reflecting the historical accumulation level of thermal excitation. If the cumulative change suddenly increases, it may mean that the brake shoe has been working at high temperatures for a long time and there is a risk of fatigue.

[0051] When any of the above-mentioned change indicators exceeds the defined normal wear tolerance range, the following condition is met: or or This indicates that if any one of the three indicators exceeds the corresponding change threshold in the j-th service stage, the structural evolution in that stage is considered abnormal, and the segment corresponding to service stage j is marked as a brake shoe ex-situ evolution marker. .

[0052] in, This represents the rate of change of the root mean square value of vibration acceleration between the j-th and (j+1)-th service stages. This represents the mileage difference between two adjacent service phases. , , These represent the threshold values ​​for abnormal changes in vibration, strain, and temperature rise, respectively. The peak strain value is at stage j. Let RMS be the vibration in stage j under condition i. This represents the change in peak strain between the j-th and (j+1)-th service stages. This represents the peak strain during the j-th service stage under the i-th operating condition. This represents the cumulative change in the slope of the temperature rise from stage 1 to stage j. This represents the temperature rise slope during the m-th service stage under the i-th operating condition.

[0053] The value range is 0.05–0.15 g / km. Based on statistical results under normal operating conditions, the root mean square value of vibration generally changes little between adjacent service stages. If the growth rate exceeds 10–20%, it indicates that there are signs of structural disturbance.

[0054] Value range: 50–100με (micro-strain). The range is determined by the fluctuation range of the stable range in the actual measured strain signal. Exceeding this value usually means that the stress area of ​​the structure has changed, including brake shoe misalignment or poor fit.

[0055] Value range: 2–5℃ / km (cumulative slope change). During the normal heating process of the brake shoe, its temperature slope change is affected by friction and environment. If the cumulative slope jumps, it is often related to problems such as friction area displacement and high temperature heat accumulation, indicating the risk of thermal fatigue.

[0056] All thresholds can be adaptively adjusted based on the specific vehicle platform, historical maintenance data, and sensor sensitivity.

[0057] The out-of-position evolution marker is a segment of change that deviates significantly from the normal wear pattern during the evolution of the brake shoe structure response over time under a certain fixed working condition. Essentially, it is the location result of an abnormal point or abnormal time period, indicating that the brake shoe has undergone a slight positional shift, loosening or change of the contact structure, or asymmetry of the force transmission path. The out-of-position evolution marker is a key marker for dynamically monitoring the health of the brake shoe structure and can be used for early warning, assisting in fault location, and guiding the determination of maintenance priorities.

[0058] S22, Lateral comparison between different operating conditions to extract preliminary damage evolution markers: The core is to compare the response characteristics under different operating conditions within the same service stage, analyze the response sensitivity of the brake shoe structure to load changes and track gradient changes, in order to identify those spatiotemporal segments where damage evolution may be occurring. Within the same service stage, for example, when the vehicle has traveled 400,000 kilometers, different operating conditions are selected for comparison: Loaded vehicle vs. empty vehicle: Represents the change in load; Slope vs. straight road: Represents the change in braking intensity.

[0059] Comparing loaded and empty vehicles, observe the change in peak strain: if the strain under loaded conditions is much greater than that under empty conditions, it indicates that the structure is very sensitive to load changes. Comparing ramps and straight sections, observe the change in the rate of temperature rise: if the rate of temperature rise increases significantly when braking on a ramp, it indicates that the structure is very sensitive to changes in braking intensity.

[0060] If the brake shoe structure is healthy, its response to different operating conditions should be linear and stable. If it becomes particularly sensitive to changes in operating conditions at a certain stage, it may be due to localized material fatigue, structural cracks or abnormal wear, or a decrease in overall stiffness. Two sensitivity coefficients are calculated to determine if the sensitivity is too high: one is the load sensitivity coefficient, which reflects the degree of strain amplification; the other is the braking strength sensitivity coefficient, which reflects the degree of thermal response enhancement. If either of these two coefficients exceeds a set threshold, the current spatiotemporal segment is marked as a preliminary brake shoe damage evolution marker.

[0061] The specific details are as follows: Under the same service stage j, different operating conditions were selected for horizontal index comparison to quantify the structural response sensitivity to load and line conditions: Load sensitivity coefficient (loaded vehicle vs. empty vehicle): This indicates the strain amplification ratio of the loaded vehicle condition to the unloaded vehicle condition under the same service stage. If this value is large, it means that the brake shoe is very sensitive to load changes and there is a risk of structural degradation or damage.

[0062] Braking intensity sensitivity coefficient (slope vs. straight road): This indicates the rate of temperature increase under the same service stage, specifically the rate of temperature rise under ramp conditions compared to straight-road conditions. A larger value indicates that the brake shoes are prone to abnormal heat load accumulation under high-intensity braking, suggesting a potential risk of thermal damage.

[0063] If the following conditions are met: or If the current structural condition is determined to be abnormally sensitive to load or line changes, it is marked as a preliminary brake shoe damage evolution marker. .

[0064] in, This represents the peak strain of a heavy-duty vehicle under the j-th service stage. This represents the peak strain of the unloaded vehicle under the j-th service stage. This represents the temperature rise slope under ramp conditions during the j-th service phase. This represents the temperature rise slope under straight road conditions during the j-th service stage. The threshold for the strain sensitivity coefficient is set at 0.25–0.35 (i.e., an amplification of 25% to 35%). Based on historical operation and maintenance data, under normal load changes, the peak strain generally fluctuates within ±15%. If the amplification exceeds 25%, it is often related to fatigue damage. The threshold for judging temperature sensitivity coefficient is set at 0.3–0.5 (i.e., 30% to 50% thermal response enhancement). Based on field tests and simulations, under healthy conditions, the difference in the heating slope between ramps and flat surfaces does not exceed 30%. If the heating rate of the ramp is much higher than this, it indicates heat accumulation or abnormal friction.

[0065] S23, as mentioned above, a batch of preliminary damage evolution markers have been extracted through sensitivity under load and braking conditions. However, these markers are mainly based on macroscopic responses to strain and temperature, and cannot fully confirm the existence of internal material damage. To verify and enhance the accuracy of the preliminary damage markers, two physical signals are introduced: acoustic emission signal energy release rate and temperature gradient distribution, to identify microcrack propagation behavior, ultimately obtaining a more reliable comprehensive brake shoe damage evolution marker. Acoustic emission signal: When cracks propagate or fracture occur inside a material, acoustic emission phenomena will occur. By using an acoustic emission sensor installed on the back plate of the brake shoe, the acoustic emission waveform signal is recorded, and the energy release rate is calculated within a specific time window to measure whether a structural rupture event has occurred.

[0066] Temperature gradient signal: Microcrack propagation often occurs in areas of localized heat concentration, so the surface temperature gradient measured by an infrared temperature sensor can be used to detect potential high-risk hotspots.

[0067] Coupled judgment model: If at a certain moment: Acoustic emission energy suddenly increases; The temperature gradient also significantly increases in the same region; Furthermore, their positions are close to each other (within the allowable error range). This indicates that microcrack propagation is likely occurring within the material in that region. This spatial-temporal coupling of dual signals is more reliable than a single feature and can significantly reduce the false positive rate.

[0068] Spatiotemporal segments that satisfy the above coupling conditions will be marked as integrated brake shoe damage evolution markers.

[0069] The specific scheme includes: enhancing damage labeling based on coupled acoustic emission and temperature gradient signals; targeting the aforementioned preliminary damage evolution labeling. We introduced acoustic emission energy release rate and temperature gradient for joint verification: Calculation of acoustic emission energy release rate: This formula represents the calculation within a time window. The average energy of the internal acoustic emission signal is used to assess whether high-energy events such as structural microcrack propagation have occurred.

[0070] Surface temperature gradient calculation (with spatial location x as reference): This formula represents the rate of temperature change at a certain location on the surface of the brake shoe, that is, the degree of temperature increase per unit length. A large temperature gradient means that there may be severe friction or heat concentration in that area, indicating potential damage.

[0071] Constructing a coupled judgment model: If in a certain time segment Within, the following conditions must be met simultaneously: and and That is, only if: Acoustic emission energy exceeded the warning value; The temperature gradient of the heat source is also abnormal; Furthermore, the two locations must be close to each other (within the tolerance range). Only when all three conditions are met simultaneously can it be considered that a real microcrack propagation behavior has occurred at this location, and the location is determined to be an internal microcrack propagation behavior, marked as a comprehensive brake shoe damage evolution marker. .

[0072] in, The instantaneous amplitude of the acoustic emission signal. The acoustic emission energy release rate over a certain time period. For energy statistics window, Let x be the temperature gradient on the surface of the brake shoe at spatial location x. Indicates the spatial location of the current temperature anomaly area. Indicates the current spatial location of the sound source. The threshold for acoustic emission energy release rate is defined as 5–20 mV² / ms or a dB-level energy index, determined based on equipment characteristics and historical experimental data. Values ​​exceeding this threshold often correspond to microcrack propagation or rapid crack tip movement. The threshold value for abnormal surface temperature gradient is 5–10℃ / cm. In infrared images, temperature differences are generally small in healthy areas. If there are rapid temperature changes in local areas, it indicates the possible presence of areas with heat accumulation / friction anomalies. The spatiotemporal matching tolerance between the sound source and the heat source is set to 1–3 cm (or equivalent sensor accuracy). The tolerance is set considering factors such as the physical size of the brake shoe, installation error, and signal delay to ensure matching of the judgment area without being too stringent.

[0073] S24, Three-dimensional coding and structural fingerprint sequence generation: The identified brake shoe displacement and damage markers need to be standardized and coded for archiving to form a structural fingerprint sequence with spatial, operating condition and temporal characteristics, which will facilitate subsequent automated identification, classification, prediction and maintenance decisions.

[0074] Two types of key structural state markers have been identified in the previous steps: Brake shoe displacement evolution marker: This indicates that the brake shoe may have experienced slight positional shifts or changes in assembly status over a certain period of time. Comprehensive brake shoe damage evolution markers: These indicate that material fatigue or microcrack propagation occurred within the brake shoe structure at certain times and locations.

[0075] These tags are just events. In order for the system to organize, identify and track them over the long term, each tag must be coded and described in a structured way. Therefore, each tag is given three dimensions of information.

[0076] Specifically, this means using the aforementioned ectopic evolution markers With integrated damage evolution markers Archived uniformly according to the following three-dimensional encoding method: Spatial coding (installation location): for example, front left, rear right; Operating condition level coding: For example, the coding for heavy vehicle-ramp is as follows: ; Service phase numbering: break-in period is 1, stable period is 2, and end period is 3; The final result is the fingerprint sequence of brake shoe displacement and damage structure required for structural defect identification: ;in, P is the installation location code, C is the operating condition level code, and j is the service stage number.

[0077] Spatial coding (P): Marks the specific installation location of the problematic brake shoe, such as the left front brake shoe or the right rear brake shoe, using a number or abbreviation to ensure accurate problem location.

[0078] Operating condition level code (C): Indicates the operating conditions under which the mark appears, such as loaded vehicle + slope, empty vehicle + flat road. Such combinations represent the complexity and load status of the operating condition and are coded in a manner similar to CRP.

[0079] Service stage number (j): indicates the current service stage of the brake shoe, such as the initial break-in stage as 1, the stable service stage as 2, and the end of wear stage as 3. This allows us to know whether the problem occurred early or evolved during the aging process.

[0080] S3, Identify structural defects and output maintenance priorities: Match the fingerprint sequence of the misplaced brake shoe and damaged structure with a pre-established structural defect pattern library, output identification results including structural defect type and severity level, and generate maintenance priorities; when a persistent low-level structural defect trend is detected, the trend is used to adaptively update the health boundary to improve the sensitivity of early defect identification.

[0081] Specifically, it includes: S31, Feature fingerprint and defect pattern matching identification: The fingerprint sequence of the valve shoe displacement and damage structure is used for identification. With a pre-established library of structural defect patterns Perform feature comparison.

[0082] Structural Defect Pattern Library Based on test bench data and typical faulty prototype vehicle data, the following typical defect feature patterns are constructed: Brake beam bending ( ); Loose pull rod ( ); boom deformation ; Crack in the support ( ); Define the fingerprint feature vector as The defect pattern vector is The structural similarity is calculated as follows: ;in, The cosine similarity between a fingerprint and a pattern of the k-th type of defect is represented by the following expression: " indicates the vector dot product operation. Given the vector magnitude, identify the item with the highest similarity. This is a type of defect in the current brake shoe structure.

[0083] S32, Structural Defect Severity Determination: Extract key features from the structural fingerprint (vibration RMS, peak strain, temperature rise slope), denoted as... , and the defect standard threshold vector Compare and calculate the degree of deviation: According to the degree of deviation Divide into categories based on the set grading thresholds: It is determined to be a serious defect; It is determined to be a moderate defect; It is determined to be a low-level defect.

[0084] in, The threshold for severity assessment is 0.4–0.5, indicating that the deviation between the current observed characteristics and a healthy structure exceeds 40%–50%, meaning that the structure has already shown significant deformation, loosening, or cracks. The threshold for determining the intermediate level is 0.2–0.3, which indicates that the structure has slightly degraded, but is still within a controllable range.

[0085] This section compares the detected brake shoe structural state (i.e., structural fingerprint) with existing typical defect samples to determine whether a certain structural defect exists and what type of defect it is. After each test, you obtain a brake shoe structural fingerprint, which is a vector composed of multiple features, such as vibration intensity, peak strain, and temperature rise slope, all of which represent the current health status of the brake shoe. A pre-established structural defect pattern library stores feature templates for different types of typical faults, such as the characteristic changes that occur when the brake beam is bent, and the response pattern that occurs when the boom is deformed. These are standard patterns summarized in advance on test benches and real prototype vehicles. The current brake shoe structural fingerprint (a set of feature data) is compared with each defect pattern in the library using cosine similarity. Essentially, this method judges the degree of similarity between two vectors (structural features) in direction. The more consistent the directions, the higher the similarity; the more skewed the directions, the less similar the two states are.

[0086] S33, based on the severity level of defects and their development trend, prioritize maintenance tasks: Emergency maintenance: The defect severity level is high, and the indicators continue to rise over multiple periods; Planned maintenance: Defect level 1, stable or slowly developing; Observation and maintenance: The defect level is low, but it recurs.

[0087] Define the current defect development trend and the structural response value of the current cycle. and the value of the previous period Take the difference, then divide by the running distance or time between these two cycles. This yields the characteristic rate of change per unit time / unit distance, expressed as: ;in, This represents the cumulative mileage or running time between two cycles. The larger this trend, the more rapidly the problem is worsening.

[0088] S34, Health Boundary Adaptive Update Mechanism: The aim is to dynamically update the health threshold of brake shoes by tracking seemingly minor but persistent defects, enabling earlier problem detection and improved identification sensitivity. Some brake shoes may exhibit low-level defects (minor deviations) in a single inspection, but if this state occurs three times consecutively (i.e., three maintenance cycles), it indicates a potential slow deterioration rather than an occasional anomaly. If this chronic degradation is not addressed promptly, it could suddenly escalate into a serious failure. To confirm the existence of this latent degradation trend, vibration growth rate, temperature change slope, and cumulative acoustic emission energy are extracted from these three cycles. These three quantities are combined into a trend feature vector, representing the brake shoe's current state of slow deterioration. Originally, there was a fixed set of health boundary thresholds, such as maximum vibration and maximum strain. Now, the trend vector is weighted and superimposed to obtain a new, adaptively adjusted threshold. If you find that certain indicators are slowly increasing but have not yet exceeded the old threshold, then lower the tolerance in advance and give an early warning. The updated threshold will be written into the brake shoe file, and all future detection and judgment will be based on this new threshold to continuously track the health status of the brake shoe.

[0089] The specific solution is as follows: When a brake shoe position shows a low-level defect trend in three consecutive maintenance cycles (e.g., j-2, j-1, j), it is considered a latent deterioration trend. At this time, the following three indicators are extracted: Vibration growth rate ; Temperature change slope ; Cumulative acoustic emission energy : ; Constructing a feature trend vector: ; Use it for adaptive updates of health boundaries: ;in, The updated coefficients represent the weighting of the trend on the threshold adjustment. The updated coefficients will then be... The structural response profile of the brake shoe under multiple operating conditions is written into the profile and used as a new criterion in subsequent testing to improve the sensitivity of early defect identification.

[0090] I. The Structural Defect Pattern Library is a collection of multi-dimensional response feature vector templates for various typical structural defects under different operating conditions and service stages, categorized by defect type. It primarily includes the following four types of typical defects: 1. Brake beam bending ( ): 1.1. Typical response characteristics: The vibration acceleration RMS increased significantly (mostly concentrated in the longitudinal component); The strain distribution is asymmetrical, and the force transmission is weakened in the later stage. The rate of temperature rise is moderate, with significant fluctuations. A faint acoustic emission signal is emitted, but the frequency is not dense.

[0091] 1.2. Spatial location characteristics: This often occurs on multiple wheel brake shoes on the same side; The strain peak value at the intermediate installation position deviated significantly.

[0092] 2. The tie rod is loose ( ): 2.1. Typical response characteristics: The vibration RMS shows intermittent jumps and the waveform is unstable; The strain peak fluctuates greatly and has poor repeatability; The temperature rise slope is normal or low; Acoustic emission signals are frequent but have low energy.

[0093] 2.2. Spatial characteristics: It occurs intermittently between multiple wheels; Combining the installation location code makes it easy to identify periodic failure points.

[0094] 3. Deformation of the suspension rod ( ): 3.1. Typical response characteristics: The peak strain is significantly higher on the off-center loading side; The vibration signal exhibits directional peak shift; The temperature response shows uneven local increases; The acoustic emission signal exhibits a stress-concentrated burst.

[0095] 3.2. Spatial characteristics: It often occurs in vehicles with a history of long-term single-sided braking or heavy load.

[0096] 4. Cracks in the support ( ): 4.1. Typical response characteristics: The vibration signal showed no significant increase. Slight fluctuations in strain values; The acoustic emission energy release rate increased sharply and the location became concentrated. The temperature slope is slightly higher, with a delayed drop.

[0097] 4.2. Multidimensional characteristics: It belongs to the structure with dual anomalies of "thermal-acoustic" but normal "force-vibration"; it is the key target of acoustic emission + temperature coupling analysis.

[0098] II. The structural defect pattern library is constructed collaboratively through the following three types of data sources: 1. Test bench simulation test: On a dedicated railway freight car braking system test bench, simulate specific defect states, including artificially adjusting loose tie rods and offset suspension rods, and conduct controlled variable tests under different load / slope / temperature environments.

[0099] Data collected includes: triaxial vibration signals, strain response, surface temperature distribution, acoustic emission signals, and control parameters such as brake cylinder pressure / acceleration.

[0100] Output the standard response template for each type of defect and the feature variation boundary under the defect threshold.

[0101] 2. Typical Fault Prototype Vehicle Operation Data: Collect prototype vehicle data from historical on-site maintenance records that clearly show typical structural faults, and analyze the multi-dimensional response evolution process before and after each defect occurs through retrospective analysis.

[0102] Data sources: on-site sensor data collection and records, train depot fault registration system, and maintenance records and photos / videos from train inspection personnel.

[0103] Processing method: The data from the three periods before the failure are taken as the pre-defect state, and the failure occurrence period is marked as the critical sample. Typical evolution curves are generated by normalization and sample clustering.

[0104] 3. To address defects with limited or rare samples, virtual response samples are generated using finite element simulation modeling to supplement the number of training samples.

[0105] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0106] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying structural defects related to misalignment and damage characteristics of railway freight car brake shoes, characterized in that, Includes the following steps: S1. During the entire life cycle of railway freight cars, based on the braking process under multiple typical working conditions, the vibration response, deformation response and temperature response of the brake shoe are collected. The braking process is divided into multiple stages and archived in layers according to working condition type and service stage to construct a multi-working-condition structural response archive of the brake shoe. S2, using the brake shoe multi-condition structural response file as input, through longitudinal comparison along the time axis within the same condition layer and lateral comparison between different condition layers, the brake shoe ex-situ evolution markers and initial brake shoe damage evolution markers that are sensitive to structural state are screened out: Among them, for the selected brake shoe damage evolution markers, a coupling analysis of acoustic emission signal energy release rate and brake shoe surface temperature gradient is introduced to identify the microcrack propagation behavior inside the brake shoe material due to long-term alternating load accumulation, and obtain a comprehensive brake shoe damage evolution marker characterizing the fatigue state of the material; all evolution markers are encoded according to spatial location, working condition level and service stage to form a fingerprint sequence of brake shoe displacement and damage structure. S3, match the fingerprint sequence of the misplaced and damaged brake shoe structure with the pre-established structural defect pattern library, output the identification results including the structural defect type and severity level, and generate maintenance priorities; when a persistent low-level structural defect trend is detected, the trend is used to adaptively update the health boundary to improve the sensitivity of early defect identification.

2. The structural defect identification method for misalignment and damage characteristics of railway freight car brake shoes according to claim 1, characterized in that, Throughout the entire life cycle of the railway freight car, vibration response is collected by accelerometers placed at the brake shoe mounting base, deformation response is collected by strain gauges placed at key stress points of the brake beam, and temperature response is collected by infrared temperature sensors placed at the brake shoe back plate. Simultaneously, brake cylinder pressure signals are collected as a benchmark for dividing working conditions.

3. The structural defect identification method for misalignment and damage characteristics of railway freight car brake shoes according to claim 2, characterized in that, Based on the characteristics of the brake cylinder pressure signal, the braking process is divided into a brake establishment stage, a pressure holding stage, and a release stage, with the pressure holding stage selected as the brake shoe action segment.

4. The structural defect identification method for misalignment and damage characteristics of railway freight car brake shoes according to claim 3, characterized in that, The vibration response, deformation response, and temperature response data are classified into first-level categories according to empty / loaded vehicle operating conditions and straight road / long slope operating conditions. Within each operating condition level, the service stages are further divided according to vehicle mileage, including the initial break-in stage, stable service stage, and end-of-wear stage. The response data of multiple brake shoe action segments under the same operating condition and the same service stage are subjected to time-domain and frequency-domain feature extraction, including the root mean square value of vibration acceleration, peak strain, and temperature rise slope, to construct a multi-condition structural response profile of brake shoes with unified feature dimensions.

5. The structural defect identification method for misalignment and damage characteristics of railway freight car brake shoes according to claim 1, characterized in that, The longitudinal comparison along the time axis within the same working condition layer includes calculating the growth rate of the root mean square value of vibration acceleration, the shift trend of the strain peak value, and the cumulative change of the temperature rise slope. Abnormal change points that exceed the normal wear range are marked as brake shoe displacement evolution markers.

6. The structural defect identification method for misalignment and damage characteristics of railway freight car brake shoes according to claim 5, characterized in that, The lateral comparison between different working conditions includes identifying the load sensitivity coefficient by the difference in strain peak value under loaded and empty vehicle conditions, identifying the braking intensity sensitivity coefficient by the difference in temperature rise slope under long slope and straight road conditions, and marking the change point of the sensitivity coefficient exceeding the threshold as the brake shoe damage evolution marker.

7. The structural defect identification method for misalignment and damage characteristics of railway freight car brake shoes according to claim 6, characterized in that, For the selected initial brake shoe damage evolution markers, acoustic emission signals collected by acoustic emission sensors arranged on the back plate of the brake shoe are introduced, and the energy release rate of the acoustic emission signals is calculated. Combined with the temperature gradient distribution on the surface of the brake shoe collected by infrared temperature sensors, a coupled analysis model of energy release rate and temperature gradient is established. When the abrupt change point of energy release rate coincides with the abnormal temperature gradient region in time and space, it is determined to be the microcrack propagation behavior inside the brake shoe material, thus obtaining a comprehensive brake shoe damage evolution marker characterizing the fatigue state of the material.

8. The structural defect identification method for misalignment and damage characteristics of railway freight car brake shoes according to claim 1, characterized in that, The S2 further includes encoding the brake shoe displacement evolution marker and the comprehensive brake shoe damage evolution marker with three-dimensional codes according to the spatial installation location, working condition level code and service stage number of the brake shoe, forming a brake shoe displacement and damage structure fingerprint sequence with spatiotemporal characteristics.

9. A structural defect identification method for railway freight car brake shoe displacement and damage characteristics according to claim 1, characterized in that, The structural defect pattern library is constructed based on test bench data and typical faulty prototype vehicle data, including typical feature patterns of brake beam bending, tie rod loosening, suspension rod deformation, and bracket cracking; by calculating the similarity between the structural fingerprint sequence and each defect pattern, the structural defect type is identified, and the severity level of the structural defect is determined according to the deviation of the abnormal feature amplitude from the standard threshold. Maintenance priorities are generated based on the severity level and development trend of structural defects. Defects with high severity level and rapid deterioration are marked as urgent maintenance, defects with medium severity level but stable are marked as planned maintenance, and low-level defects that persist are marked as observation maintenance.

10. A structural defect identification method for misalignment and damage characteristics of railway freight car brake shoes according to claim 9, characterized in that, The S3 further includes: when the same brake shoe location shows a low-level structural defect trend in three consecutive maintenance cycles, extracting the characteristic parameters of the trend, including vibration growth rate, temperature change slope and acoustic emission energy accumulation, for adaptively updating the health boundary threshold, and writing the updated health boundary into the brake shoe multi-condition structural response file.

Images