Modular and connectable track inspection vehicle: multi-device collaborative data acquisition method and system

By using a modular track inspection vehicle for unified timing, dynamic synchronization compensation, and weighted fusion, the problems of low inspection efficiency, poor accuracy, and inaccurate risk assessment in existing technologies have been solved, enabling efficient and accurate track inspection and maintenance decisions.

CN123104735APending Publication Date: 2026-11-17CHINA RAILWAY FIRST GROUP CO LTD +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611222086.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-12
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing track inspection technologies suffer from problems such as separation of operations, inability to directly correlate data, low time synchronization accuracy, inaccurate evaluation methods, and incomplete risk assessment, resulting in low inspection efficiency, high costs, and underestimated risks.

Method used

The modular and connectable track inspection vehicle integrates image acquisition, profile point cloud acquisition, positioning and attitude determination, and mileage acquisition modules. Through the synchronization control module, it realizes unified time synchronization, dynamic time synchronization compensation, weighted fusion, and comprehensive defect risk assessment, forming a digital archive of track status under a unified time scale.

Benefits of technology

It enables simultaneous detection by multiple devices, improving detection efficiency by more than 50%, enhancing time synchronization accuracy and data fusion accuracy, accurately assessing track profile deterioration trends, providing quantitative maintenance decision-making basis, and reducing reliance on manual handling.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123104735A_ABST
    Figure CN123104735A_ABST
Patent Text Reader

Abstract

This invention discloses a modular, connectable track inspection vehicle multi-device collaborative data acquisition method and system, relating to the field of rail transit inspection technology. The method includes: a synchronization control module generating trigger signals based on external time synchronization and mileage pulses, distributing them to an image and profile point cloud acquisition module, which responds to the acquisition and feeds back the original timestamp. The module dynamically compensates for the acquisition time based on running speed and attitude angular velocity to obtain a synchronization compensation timestamp, and then weights and fuses the original data to obtain fused image and point cloud data. Based on the fused point cloud, the accumulated track wear is calculated; based on the fused image, a damage score is obtained; and combined with the track gauge deviation extracted from the point cloud, a comprehensive track defect risk index is calculated. After comparing this index with a preset threshold, a maintenance decision is output.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of rail transit detection, and particularly relates to a modular and spliceable multi-device cooperative acquisition method and system for a track detection vehicle. Background Art

[0002] Track inspection is a critical maintenance link in rail transit systems such as railways and subways, whose core purpose is to ensure the safety of track structures and normal operation of equipment through systematic inspection, so as to prevent train operation accidents. Long-term repeated vibration during train operation will gradually cause bolt loosening and elastic bar fatigue fracture; long-term load of trains will lead to gradual settlement and deformation of the track bed, causing changes in track geometric dimensions; wheel-rail creepage during train starting and braking will further aggravate the deformation of the rail profile; after the fatigue stress generated by repeated train loads exceeds the strength of the steel, cracks will form and expand into internal injuries. If not detected in time, the cracks will expand into rail head internal injuries, which in severe cases will directly lead to rail breakage and derailment accidents; fastener failure will cause rail creeping and gauge overrun; cracking and failure of sleepers will reduce the track bearing capacity. All the above defects directly threaten the operation safety, and need to be detected and handled in time through regular inspection.

[0003] In the existing detection technologies, track inspection mainly relies on manual track patrolling, using tools to inspect defects such as rail surface damage, abnormal fasteners, sleeper cracking, track bed cracking, and foreign matters on the track bed; for detection items such as fastener state, track geometric parameters, and rail profile, spot checks are mostly carried out through torque wrenches, manual track gauges, strings, and handheld profilometers. Due to the short time of the skylight period when the operation is suspended at night, the manual detection method has high labor intensity, large subjective error and low efficiency, and the detection data depends on human subjective judgment, which is difficult to quantify and trace. There also exists a method of manually pushing a single rail surface state detection device for detection, but this method has a single detection content and a low manual pushing speed, cannot complete multiple detection items within one skylight period, and cannot make full use of the valuable skylight period to conduct a comprehensive detection of the track.

[0004] Existing technologies have the following objective drawbacks: First, the operations are separated; track defect detection and rail profile detection need to be carried out separately, increasing the detection cycle and operating costs. Second, data cannot be directly correlated; due to the separate operations, data collected by different devices cannot be aligned in time and space, and image data cannot be fused with profile point cloud data, resulting in a large amount of manpower required for data registration later, leading to long processing cycles and high costs. Third, existing time synchronization methods typically only use fixed pulse triggers, without considering the impact of changes in the inspection vehicle's speed and vehicle vibration on the actual response times of each sensor. The resulting dynamic offset leads to a significant decrease in the matching accuracy between image data and point cloud data in areas where the detection vehicle changes speed or the road is bumpy; fourth, existing methods for evaluating profile wear mostly use a single average or root mean square value for statistical analysis, which fails to comprehensively reflect the cumulative fatigue effect caused by local abrupt damage and long-term repeated loading, making it difficult to accurately assess the deterioration trend of the rail profile; fifth, existing methods for assessing defects usually output the image damage identification results and the geometric parameter over-limit judgment separately, without establishing a coupling relationship between image features and profile features, which easily leads to the omission of defect sections where the risk is underestimated under the combined effect of image damage and profile wear. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a modular, connectable track inspection vehicle with multi-device collaborative data acquisition method. The method is applied to a track inspection vehicle, which includes a modular vehicle body. The modular vehicle body is equipped with an image acquisition module, a profile point cloud acquisition module, a positioning and attitude determination module, a mileage acquisition module, and a synchronization control module, comprising: S1: The synchronization control module completes the timing according to the external timing signal output by the positioning and attitude determination module, and generates a trigger signal according to the pulse signal output by the mileage acquisition module, and distributes the trigger signal to the image acquisition module and the profile point cloud acquisition module to obtain the timing-accompanied trigger signal; S2: The image acquisition module and the profile point cloud acquisition module respectively respond to the timed trigger signal to acquire data, obtain the original image data, the original point cloud data and their respective original timestamps, and feed the original timestamps back to the synchronization control module; S3: The synchronization control module dynamically compensates the acquisition time of the image acquisition module and the profile point cloud acquisition module based on the original timestamp, the running speed information output by the mileage acquisition module, and the attitude angular velocity information output by the positioning and attitude determination module, to obtain the synchronization compensation timestamp. S4: The synchronization control module performs weighted fusion of the original image data and the original point cloud data based on the synchronization compensation timestamp to obtain fused image data and fused point cloud data; S5: The synchronization control module calculates the cumulative wear index of the track profile based on the fused point cloud data; S6: The synchronization control module obtains an image damage score based on the fused image data, and calculates a comprehensive track defect risk index by combining the cumulative wear index and the track gauge deviation extracted from the fused point cloud data. S7: The synchronous control module compares the track comprehensive defect risk index with the preset risk threshold and outputs maintenance decision information.

[0006] Furthermore, the synchronization compensation timestamps include: , in, For the first Each sensor module at time Synchronization compensation timestamp, For the first The initial static timescale drift constant of each sensor module For the first Thermal drift attenuation coefficient of each sensor module For the first The velocity coupling drift coefficient of each sensor module For the mileage acquisition module at time The measured speed of the inspection vehicle For the first The speed nonlinearity sensitivity index of each sensor module For the inertial measurement unit in the positioning and attitude determination module at time... The measured angular velocity of the test vehicle attitude, For the first The initial phase angle of each sensor module.

[0007] Furthermore, step S4 also includes: calculating normalized fusion weights and performing weighted fusion on the original image data and the original point cloud data; The calculation of normalized fusion weights includes: , in, For the first The sensor module in the first Synchronization Compensation Timestamp Normalized fusion weights at the location, This represents the total number of sensor modules participating in this fusion. For the first The sensor module in the first Synchronization Compensation Timestamp Signal quality indicators of the collected data The nonlinear amplification index represents the signal quality. This is the vibration suppression coefficient. For the first The sensor module is in the first Synchronization Compensation Timestamp Within the centered acquisition time window, by The calculated standard deviation of angular jitter, For the first The sensor module in the first Synchronization Compensation Timestamp Signal quality indicators of the collected data For the first The sensor module is in the first Synchronization Compensation Timestamp Within the centered acquisition time window, by The calculated standard deviation of angular jitter.

[0008] Furthermore, the cumulative wear index of the track profile was calculated, including: , in, The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. neighborhood The coordinates of the nearest location within, The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates First The profile deviation characteristic quantity that passed the second test. For the order parameter of the spatially weighted power average, for right Normalized space weights, For the recursive forgetting factor that passes the cross-detection, This is the fatigue accumulation coefficient. This is the fatigue cumulative nonlinear index.

[0009] Furthermore, calculate the mileage coordinates along the direction of track extension. neighborhood Neighboring location coordinates First Profile deviation characteristic quantity passed in the second test include: , in, It is the arc length along the profile of the rail cross-section. The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates The total arc length of the profile of the rail cross-section measured at the location. The first point cloud data obtained from the fused point cloud data The second detection uses the coordinates of the mileage position along the direction of track extension. neighborhood Neighboring location coordinates Arc length The local deviation between the measured profile of the rail at the location and the preset design reference profile. This is the order parameter.

[0010] Further calculations right Normalized space weights ,include: , in, For and A Gaussian kernel function with parameters. The spatial bandwidth of the Gaussian kernel function. For and A Gaussian kernel function with parameters. The coordinates of the mileage position along the direction of the track extension. neighborhood The coordinates of the nearest location within.

[0011] Furthermore, the initial conditions for the recursion are: , in, The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates First The profile deviation feature quantity that passed the test.

[0012] Furthermore, the calculated comprehensive track defect risk index includes: , in, The coordinates of the mileage position along the direction of the track extension. The comprehensive defect risk index of the track. The coordinates of the mileage position along the direction of the track extension. Image damage score at the location, This serves as a normalized reference value for image damage scoring. The coordinates of the mileage position along the direction of the track extension. The cumulative wear index at the location, This serves as a normalized reference value for the cumulative wear index. The mileage coordinates extracted from the fused point cloud data along the track extension direction. The deviation between the measured track gauge and the preset standard track gauge. This serves as a normalized reference value for track gauge deviation. Image damage weights, For the weight of profile wear, As the weight for track gauge deviation, The order parameter of the weighted power average. The coordinates of the mileage position along the direction of the track extension. The cross-coupling feature between image damage score and cumulative wear index. The upper limit coefficient for saturation. It is a semi-saturated constant. It is a synergy index.

[0013] Furthermore, calculate the mileage coordinates along the direction of track extension. Cross-coupling features of image damage score and cumulative wear index include: .

[0014] This invention also proposes a modular, connectable track inspection vehicle multi-device collaborative data acquisition system. The system is applied to a track inspection vehicle, which includes a modular vehicle body. The modular vehicle body is equipped with an image acquisition module, a profile point cloud acquisition module, a positioning and attitude determination module, a mileage acquisition module, and a synchronization control module, comprising: The timing module is used by the synchronization control module to complete the timing according to the external timing signal output by the positioning and attitude determination module, and to generate a trigger signal according to the pulse signal output by the mileage acquisition module, and distribute the trigger signal to the image acquisition module and the profile point cloud acquisition module to obtain the timing trigger signal; The acquisition module is used for the image acquisition module and the contour point cloud acquisition module to respectively respond to the timed trigger signal to acquire data, obtain raw image data, raw point cloud data and their respective original timestamps, and feed the original timestamps back to the synchronization control module; The synchronization compensation module is used by the synchronization control module to dynamically compensate the acquisition time of the image acquisition module and the profile point cloud acquisition module based on the original timestamp, the running speed information output by the mileage acquisition module and the attitude angular velocity information output by the positioning and attitude determination module, so as to obtain the synchronization compensation timestamp. The fusion module is used by the synchronization control module to perform weighted fusion of the original image data and the original point cloud data based on the synchronization compensation timestamp to obtain fused image data and fused point cloud data. The cumulative wear index calculation module is used by the synchronization control module to calculate the cumulative wear index of the track profile based on the fused point cloud data. The track comprehensive defect risk index calculation module is used by the synchronization control module to calculate the track comprehensive defect risk index by combining the image damage score obtained from the fused image data with the cumulative wear index and the track gauge deviation extracted from the fused point cloud data. The maintenance decision information output module is used by the synchronous control module to compare the track comprehensive defect risk index with a preset risk threshold and output maintenance decision information.

[0015] Compared with the prior art, the present invention has the following advantages and technical effects: Firstly, through unified timing and trigger distribution in steps S1 to S2, the image acquisition module and the profile point cloud acquisition module can simultaneously complete the acquisition of rail surface image data, track fastener and sleeper ballast bed status data, track internal geometric parameter data and rail profile data in the same online operation, without the need for separate operations, shortening the inspection cycle and making full use of the short nighttime window.

[0016] Secondly, through the dynamic time synchronization compensation described in step S3, the dynamic influence of the vehicle's running speed and attitude angular velocity on the actual response time of each sensor module is further taken into account on the basis of the fixed period of the trigger pulse. This significantly improves the time synchronization accuracy of the vehicle on variable speed or bumpy road sections, providing a more accurate time reference for subsequent data fusion.

[0017] Third, through the weighted fusion described in step S4, the weight ratio of each sensor module data in the fusion result is dynamically adjusted according to the signal quality index and the standard deviation of angular jitter caused by vibration. This enables the fused data to have the ability to adapt to local signal quality fluctuations while ensuring accuracy, and realizes accurate matching and fusion of image data and point cloud data. Compared with the traditional step-by-step detection method, the work efficiency can be improved by more than 50%.

[0018] Fourth, by calculating the cumulative wear index as described in step S5, the generalized power norm replaces the traditional single average or root mean square value statistical method to reflect the local abrupt damage of a single detection. The spatial correlation trend of wear state between adjacent mileage positions is captured by Gaussian kernel weighted spatial smoothing in the mileage direction. Furthermore, the fatigue accumulation amplification effect is superimposed by the recursive forgetting fusion of cross-detection sequence, so that the cumulative nonlinear wear index can reflect both local abrupt damage and the spatial propagation trend of wear state and the fatigue accumulation effect caused by long-term repeated loading, thereby improving the accuracy of track profile deterioration trend assessment.

[0019] Fifth, through the calculation of the comprehensive track defect risk index described in step S6, the image damage score, profile wear index, and gauge deviation are normalized and then weighted by power average to obtain the basic risk level. A saturated rational function risk upgrade term is introduced for the cross-coupled features of image damage and profile wear. When both image damage and profile wear are significant, the basic risk level is upgraded nonlinearly with a controllable amplitude. This avoids the risk underestimation problem caused by the separate evaluation of image features and profile features in the prior art. Moreover, the saturation upper limit and half-saturation constant of the risk upgrade term are explicit parameters that can be set independently, which facilitates the targeted calibration of the risk assessment model based on on-site defect data. This provides a more comprehensive, quantitative, and adjustable basis for track maintenance decisions.

[0020] Sixth, through the associated storage described in step S7, a digital archive of track status under a unified time scale is formed. The data format is standardized and uniform, which facilitates subsequent trend analysis, problem review and accountability, and reduces the reliance on manual data processing. Attached Figure Description

[0021] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of the method in Embodiment 1 of the present invention; Figure 2 This is a system structure diagram of Embodiment 2 of the present invention; Figure 3 This is a system structure diagram of Embodiment 4 of the present invention; Figure 4This is a comparison chart of the nonlinear dynamic time synchronization compensation effect of the present invention. Detailed Implementation

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0024] The modular, connectable track inspection vehicle provided by this invention includes a vehicle body module, a sensor module, and a synchronization control module. The vehicle body module consists of several standardized splicing units. Adjacent splicing units are mechanically connected via a snap-fit ​​structure and electrically connected via an industrial Ethernet bus laid along the vehicle body. Each splicing unit has a reserved interface for installing sensor modules, allowing for the addition, removal, and replacement of sensor modules as needed for the inspection task. Drive units are located at both ends or in the middle of the vehicle body module to drive the inspection vehicle along the track. An onboard battery is installed on the vehicle body module to power the sensor modules and the synchronization control module.

[0025] The sensor module includes an image acquisition module, a profile point cloud acquisition module, a positioning and attitude determination module, and a mileage acquisition module. The image acquisition module uses a line-scan camera sensor to acquire image data of the rail surface. The profile point cloud acquisition module uses a high-precision line-structured light laser sensor to acquire point cloud data corresponding to the rail fastener status, sleeper track bed status, internal track geometric parameters, and rail profile. The positioning and attitude determination module includes a satellite positioning unit and an inertial measurement unit. The satellite positioning unit receives satellite positioning signals and outputs a second pulse timing signal. The inertial measurement unit includes a gyroscope to output the attitude angular velocity information of the inspection vehicle. The mileage acquisition module uses an encoder to output pulse signals corresponding to the mileage traveled by the inspection vehicle and can calculate the vehicle's speed information based on the time interval between adjacent pulses. The synchronization control module is connected to the image acquisition module, profile point cloud acquisition module, positioning and attitude determination module, and mileage acquisition module via the industrial Ethernet bus. It performs data processing tasks such as trigger signal distribution, time synchronization compensation, multi-source data fusion, profile wear calculation, and defect risk assessment, and stores the processing results in the industrial control computer storage unit located within the synchronization control module.

[0026] Example 1 like Figure 1As shown, this embodiment proposes a modular, connectable track inspection vehicle multi-device collaborative data acquisition method. The method is applied to a track inspection vehicle, which includes a modular vehicle body. The modular vehicle body is equipped with an image acquisition module, a profile point cloud acquisition module, a positioning and attitude determination module, a mileage acquisition module, and a synchronization control module, including: S1: The synchronization control module completes the timing according to the external timing signal output by the positioning and attitude determination module, and generates a trigger signal according to the pulse signal output by the mileage acquisition module, and distributes the trigger signal to the image acquisition module and the profile point cloud acquisition module to obtain the timing-accompanied trigger signal; Preferably, the synchronization control module uses the second pulse signal received by the satellite positioning unit in the positioning and attitude determination module to align the internal timing reference of the synchronization control module with the external satellite time reference, thereby completing the time synchronization; the synchronization control module further generates a trigger signal based on the pulse signal output by the mileage acquisition module, and distributes the trigger signal to the image acquisition module and the profile point cloud acquisition module to obtain the time-synchronized trigger signal.

[0027] S2: The image acquisition module and the profile point cloud acquisition module respectively respond to the timed trigger signal to acquire data, obtain the original image data, the original point cloud data and their respective original timestamps, and feed the original timestamps back to the synchronization control module; S3: The synchronization control module dynamically compensates the acquisition times of the image acquisition module and the profile point cloud acquisition module based on the original timestamp, the running speed information output by the mileage acquisition module, and the attitude angular velocity information output by the positioning and attitude determination module, to obtain a synchronization compensation timestamp, the effect of which is as follows: Figure 4 As shown; Specifically, for the image acquisition module and the profile point cloud acquisition module, the first... One sensor module ( ,in This indicates the image acquisition module. (Representing the profile point cloud acquisition module), the synchronization compensation timestamps obtained include: , in, For the first Each sensor module at time Synchronization compensation timestamp, For the first The initial static timescale drift constant of each sensor module is given by the factory calibration value when the detection vehicle is powered on, and the unit is seconds (s). For the first The thermal drift decay coefficient of each sensor module is used to characterize the process by which the sensor module gradually reaches thermal equilibrium as the operating time increases, and the unit is negative first power of seconds. ), For the first The velocity coupling drift coefficient of each sensor module is used to characterize the degree of influence of the vehicle's operating speed on the sensor module's response delay, and the unit is 1. , For the mileage acquisition module at time The measured speed of the inspection vehicle For the first The velocity nonlinear sensitivity index of each sensor module is a dimensionless quantity with a value range of [value missing]. , For the inertial measurement unit in the positioning and attitude determination module at time... The measured angular velocity of the test vehicle attitude, For the first The initial phase angle of each sensor module.

[0028] Example of parameter values: , (Corresponding unit) ), , ; , (Corresponding unit) ), , .

[0029] S4: The synchronization control module performs weighted fusion of the original image data and the original point cloud data based on the synchronization compensation timestamp to obtain fused image data and fused point cloud data; Preferably, the synchronization control module aligns the original image data and the original point cloud data obtained in step S2 to a unified synchronization timestamp sequence based on the synchronization compensation timestamp obtained in step S3, and assigns nonlinear fusion weights to the data of the image acquisition module and the profile point cloud acquisition module at each synchronization timestamp before performing weighted fusion to obtain fused image data and fused point cloud data (hereinafter collectively referred to as fused data).

[0030] Specifically, step S4 further includes: calculating normalized fusion weights and performing weighted fusion on the original image data and the original point cloud data; set up ( The result obtained after compensation in step S3 is the first... A synchronized timestamp The total number of synchronized timestamps, the first One sensor module ( Define the same steps as in step S3) for synchronizing timestamps. The normalized fusion weights at each location include: , in, For the first The sensor module in the first Synchronization Compensation Timestamp Normalized fusion weights at the location, This represents the total number of sensor modules participating in this fusion. For the first The sensor module in the first Synchronization Compensation Timestamp The signal quality indicators for the acquired data are as follows: for the image acquisition module, the image sharpness evaluation value is taken; for the profile point cloud acquisition module, the point cloud reflectance intensity uniformity ratio is taken. These are dimensionless quantities with a value range of [missing value]. , The nonlinear amplification index represents the signal quality. The vibration suppression coefficient is used to suppress the weighting of data with large positional errors caused by vehicle vibration during fusion, expressed in negative square radians. ), For the first The sensor module is in the first Synchronization Compensation Timestamp Within the centered acquisition time window, by The calculated standard deviation of angular jitter, For the first The sensor module in the first Synchronization Compensation Timestamp Signal quality indicators of the collected data For the first The sensor module is in the first Synchronization Compensation Timestamp Within the centered acquisition time window, by The calculated standard deviation of angular jitter.

[0031] Example of parameter values: , Image acquisition module , Profile point cloud acquisition module , .

[0032] S5: The synchronization control module calculates the cumulative wear index of the track profile based on the fused point cloud data; Preferably, the synchronization control module extracts the track's position at the mileage location based on the fused point cloud data obtained in step S4. The local deviation between the rail cross-sectional profile point cloud at the location and its adjacent mileage and the preset rail design reference profile, combined with the historical inspection records at that mileage location, is used to calculate the cumulative nonlinear wear index of the track profile through a combination of spatial kernel weighted smoothing and recursive accumulation across inspection batches.

[0033] Specifically, the cumulative wear indicators for the track profile calculated include: , in, The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. neighborhood The coordinates of the nearest location within, The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates First The profile deviation characteristic quantity that passed the second test. For the order parameter of the spatially weighted power average, for right Normalized space weights, For the recursive forgetting factor that passes the cross-detection, This is the fatigue accumulation coefficient. This is the fatigue cumulative nonlinear index.

[0034] Specifically, for the first The second test passed ( , Mileage location The total number of times the test passes is recorded cumulatively (a dimensionless integer), and the mileage coordinates along the track extension direction are calculated. neighborhood Neighboring location coordinates First Profile deviation characteristic quantity passed in the second test include: , in, The value is the arc length on the profile line of the rail cross-section, and its range is... , The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates The total arc length of the profile of the rail cross-section measured at the location. The first point cloud data obtained from the fused point cloud data The second detection uses the coordinates of the mileage position along the direction of track extension. neighborhood Neighboring location coordinates Arc length The local deviation between the measured profile of the rail at the location and the preset design reference profile. This is the order parameter.

[0035] Specifically, calculation right Normalized space weights ,include: , in, For and A Gaussian kernel function with parameters. The spatial bandwidth of the Gaussian kernel function. For and A Gaussian kernel function with parameters. The coordinates of the mileage position along the direction of the track extension. neighborhood The coordinates of the nearest location within.

[0036] Specifically, the initial conditions for the recursion are: , in, The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates First The profile deviation feature quantity that passed the test.

[0037] Example of parameter values: , , , , , , , .

[0038] Preferably, the cumulative wear index of the track profile is taken as the recursive result after the most recent successful inspection, i.e. When the mileage location Located near the start or end of the detection line, or in the vicinity When the number of available nearby mileage locations is insufficient, only for Sum the actual neighboring mileage locations and normalize them again according to the above formula. No additional neighboring points are added.

[0039] S6: The synchronization control module obtains an image damage score based on the fused image data, and calculates a comprehensive track defect risk index by combining the cumulative wear index and the track gauge deviation extracted from the fused point cloud data. Preferably, the synchronization control module obtains the mileage position based on the fused image data obtained in step S4 using a preset surface damage recognition algorithm. The image damage score is obtained by using existing technologies in image processing or deep learning. For example, a target detection model or semantic segmentation model based on convolutional neural networks, including Faster R-CNN detection model, YOLO series detection model, U-Net segmentation model, and Mask R-CNN instance segmentation model, can be used to identify cracks, spalling, scratches, and corrosion damage areas in the rail surface image. The image damage score is then calculated by weighting the proportion of identified damage pixels with the detection confidence level. Alternatively, traditional image processing methods based on gray-level co-occurrence matrix texture features combined with support vector machine classifiers, or Canny edge detection combined with morphological processing, can be used for damage area identification and scoring. This invention does not limit the specific implementation of the surface damage identification algorithm; any existing image recognition algorithm capable of outputting a quantified damage score can be used to obtain the image damage score in step S6.

[0040] Specifically, the calculated comprehensive track defect risk index includes: , in, The coordinates of the mileage position along the direction of the track extension. The comprehensive defect risk index of the track. The coordinates of the mileage position along the direction of the track extension. Image damage score at the location, This serves as a normalized reference value for image damage scoring. The coordinates of the mileage position along the direction of the track extension. The cumulative wear index at the location, This serves as a normalized reference value for the cumulative wear index. The mileage coordinates extracted from the fused point cloud data along the track extension direction. The deviation between the measured track gauge and the preset standard track gauge. This serves as a normalized reference value for track gauge deviation. Image damage weights, For the weight of profile wear, As the weight for track gauge deviation, The order parameter of the weighted power average. The coordinates of the mileage position along the direction of the track extension. The cross-coupling feature between image damage score and cumulative wear index. The upper limit coefficient for saturation. It is a semi-saturated constant. It is a synergy index.

[0041] Example of parameter values: , , , , , , , , , .

[0042] Specifically, calculate the mileage coordinates along the direction of the track extension. Cross-coupling features of image damage score and cumulative wear index include: .

[0043] S7: The synchronous control module compares the track comprehensive defect risk index with the preset risk threshold and outputs maintenance decision information.

[0044] Preferably, the synchronization control module compares the track comprehensive defect risk index obtained in step S6 with a preset risk threshold. If the track comprehensive defect risk index is not less than the risk threshold, it outputs maintenance decision information for the corresponding mileage position. The synchronization control module stores the synchronization compensation timestamp, the fused data, the cumulative wear index, the track comprehensive defect risk index, and the maintenance decision information obtained in each step in association according to mileage position and detection time, forming a traceable digital archive of track status.

[0045] As an example, the present invention can preset two levels of judgment criteria: high-risk threshold and medium-risk threshold, and combine the component that contributes the most to the comprehensive track defect risk index among the image damage score, the cumulative wear index and the track gauge deviation to generate corresponding maintenance decision information. If the comprehensive track defect risk index at a certain mileage location reaches the high-risk threshold, the output maintenance decision information can be: "The comprehensive defect risk level at this mileage location is high-risk. It is recommended to immediately limit speed and arrange manual review within 24 hours." If the comprehensive track defect risk index reaches the medium-risk threshold and the cumulative nonlinear wear index is the main contributing component, the output maintenance decision information can be: "The profile wear at this mileage location is relatively prominent. It is recommended to include it in the rail grinding plan for the next daily window." If the comprehensive track defect risk index reaches the medium-risk threshold and the gauge deviation is the main contributing component, the output maintenance decision information can be: "The gauge deviation at this mileage location exceeds the normal range. It is recommended to arrange fastener inspection and gauge adjustment operations." If the comprehensive track defect risk index is lower than the medium-risk threshold, the output maintenance decision information can be: "The condition at this mileage location is normal. It should be included in the regular inspection cycle."

[0046] Example 2 like Figure 2 As shown, this embodiment proposes a modular, connectable track inspection vehicle multi-device collaborative data acquisition system. The system is applied to a track inspection vehicle, which includes a modular vehicle body. The modular vehicle body is equipped with an image acquisition module, a profile point cloud acquisition module, a positioning and attitude determination module, a mileage acquisition module, and a synchronization control module, comprising: The timing module is used by the synchronization control module to complete the timing according to the external timing signal output by the positioning and attitude determination module, and to generate a trigger signal according to the pulse signal output by the mileage acquisition module, and distribute the trigger signal to the image acquisition module and the profile point cloud acquisition module to obtain the timing trigger signal; The acquisition module is used for the image acquisition module and the contour point cloud acquisition module to respectively respond to the timed trigger signal to acquire data, obtain raw image data, raw point cloud data and their respective original timestamps, and feed the original timestamps back to the synchronization control module; The synchronization compensation module is used by the synchronization control module to dynamically compensate the acquisition time of the image acquisition module and the profile point cloud acquisition module based on the original timestamp, the running speed information output by the mileage acquisition module and the attitude angular velocity information output by the positioning and attitude determination module, so as to obtain the synchronization compensation timestamp. The fusion module is used by the synchronization control module to perform weighted fusion of the original image data and the original point cloud data based on the synchronization compensation timestamp to obtain fused image data and fused point cloud data. The cumulative wear index calculation module is used by the synchronization control module to calculate the cumulative wear index of the track profile based on the fused point cloud data. The track comprehensive defect risk index calculation module is used by the synchronization control module to calculate the track comprehensive defect risk index by combining the image damage score obtained from the fused image data with the cumulative wear index and the track gauge deviation extracted from the fused point cloud data. The maintenance decision information output module is used by the synchronous control module to compare the track comprehensive defect risk index with a preset risk threshold and output maintenance decision information.

[0047] Since Example 2 is based on Example 1, it will not be described again.

[0048] Example 3 This invention also proposes a storage medium storing multiple instructions, which are used to implement the modular and connectable multi-device collaborative data acquisition method for track inspection vehicles.

[0049] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0050] Optionally, in this embodiment, the storage medium is configured to store program code for performing the method steps of Embodiment 1.

[0051] Example 4 This invention also proposes an electronic device, such as... Figure 3 As shown, it includes a processor and a storage medium connected to the processor. The storage medium stores multiple instructions, which can be loaded and executed by the processor to enable the processor to execute the modular and connectable track inspection vehicle multi-device collaborative acquisition method.

[0052] Specifically, the electronic device in this embodiment can be a computer terminal, which may include one or more processors and a storage medium.

[0053] The storage medium can be used to store software programs and modules, such as the modular and modular track inspection vehicle multi-device collaborative data acquisition method in this embodiment of the invention. The corresponding program instructions / modules are executed by the processor through the software programs and modules stored in the storage medium, thereby performing various functional applications and data processing, thus realizing the aforementioned modular and modular track inspection vehicle multi-device collaborative data acquisition method. The storage medium may include high-speed random access storage media, and may also include non-volatile storage media, such as one or more magnetic storage systems, flash memory, or other non-volatile solid-state storage media. In some instances, the storage medium may further include storage media remotely configured relative to the processor, which can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0054] The processor can execute the method steps of Embodiment 1 by calling the information and application stored in the storage medium through the transmission system.

[0055] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0056] In the several embodiments provided by this invention, it should be understood that the disclosed technical content can be implemented in other ways. The system embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection between units or modules, and may be electrical or other forms.

[0057] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0058] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0059] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, optical disks, and other media capable of storing program code.

[0060] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for collaborative data acquisition using multiple devices on a modular, connectable track inspection vehicle, the method being applied to a track inspection vehicle comprising a modular body, the modular body being equipped with an image acquisition module, a profile point cloud acquisition module, a positioning and attitude determination module, a mileage acquisition module, and a synchronization control module, characterized in that, include: S1: The synchronization control module completes the timing according to the external timing signal output by the positioning and attitude determination module, and generates a trigger signal according to the pulse signal output by the mileage acquisition module, and distributes the trigger signal to the image acquisition module and the profile point cloud acquisition module to obtain the timing-accompanied trigger signal; S2: The image acquisition module and the profile point cloud acquisition module respectively respond to the timed trigger signal to acquire data, obtain the original image data, the original point cloud data and their respective original timestamps, and feed the original timestamps back to the synchronization control module; S3: The synchronization control module dynamically compensates the acquisition time of the image acquisition module and the profile point cloud acquisition module based on the original timestamp, the running speed information output by the mileage acquisition module, and the attitude angular velocity information output by the positioning and attitude determination module, to obtain the synchronization compensation timestamp. S4: The synchronization control module performs weighted fusion of the original image data and the original point cloud data based on the synchronization compensation timestamp to obtain fused image data and fused point cloud data; S5: The synchronization control module calculates the cumulative wear index of the track profile based on the fused point cloud data; S6: The synchronization control module obtains an image damage score based on the fused image data, and calculates a comprehensive track defect risk index by combining the cumulative wear index and the track gauge deviation extracted from the fused point cloud data. S7: The synchronous control module compares the track comprehensive defect risk index with the preset risk threshold and outputs maintenance decision information.

2. The modular and connectable track inspection vehicle multi-device collaborative data acquisition method as described in claim 1, characterized in that, The synchronization compensation timestamps include: , in, For the first Each sensor module at time Synchronization compensation timestamp, For the first The initial static timescale drift constant of each sensor module For the first Thermal drift attenuation coefficient of each sensor module For the first The velocity coupling drift coefficient of each sensor module For the mileage acquisition module at time The measured speed of the inspection vehicle For the first The speed nonlinearity sensitivity index of each sensor module For the inertial measurement unit in the positioning and attitude determination module at time... The measured angular velocity of the test vehicle attitude, For the first The initial phase angle of each sensor module.

3. The modular and connectable track inspection vehicle multi-device collaborative data acquisition method as described in claim 2, characterized in that, Step S4 further includes: calculating normalized fusion weights and performing weighted fusion on the original image data and the original point cloud data; The calculation of normalized fusion weights includes: , in, For the first The sensor module in the first Synchronization Compensation Timestamp Normalized fusion weights at the location, This represents the total number of sensor modules participating in this fusion. For the first The sensor module in the first Synchronization Compensation Timestamp Signal quality indicators of the collected data The nonlinear amplification index represents the signal quality. This is the vibration suppression coefficient. For the first The sensor module is in the first Synchronization Compensation Timestamp Within the centered acquisition time window, by The calculated standard deviation of angular jitter, For the first The sensor module in the first Synchronization Compensation Timestamp Signal quality indicators of the collected data For the first The sensor module is in the first Synchronization Compensation Timestamp Within the centered acquisition time window, by The calculated standard deviation of angular jitter.

4. The modular and connectable track inspection vehicle multi-device collaborative data acquisition method as described in claim 1, characterized in that, The cumulative wear indexes of the track profile calculated include: , in, The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. neighborhood The coordinates of the nearest location within, The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates First The profile deviation characteristic quantity that passed the second test. For the order parameter of the spatially weighted power average, for right Normalized space weights, For the recursive forgetting factor that passes the cross-detection, This is the fatigue accumulation coefficient. This is the fatigue cumulative nonlinear index.

5. The modular and connectable track inspection vehicle multi-device collaborative data acquisition method as described in claim 4, characterized in that, Calculate the mileage coordinates along the track extension direction. neighborhood Neighboring location coordinates First Profile deviation characteristic quantity passed in the second test include: , in, It is the arc length along the profile of the rail cross-section. The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates The total arc length of the profile of the rail cross-section measured at the location. The first point cloud data obtained from the fused point cloud data The second detection uses the coordinates of the mileage position along the direction of track extension. neighborhood Neighboring location coordinates Arc length The local deviation between the measured profile of the rail at the location and the preset design reference profile. This is the order parameter.

6. The modular and connectable track inspection vehicle multi-device collaborative data acquisition method as described in claim 4, characterized in that, calculate right Normalized space weights ,include: , in, For and A Gaussian kernel function with parameters. The spatial bandwidth of the Gaussian kernel function. For and A Gaussian kernel function with parameters. The coordinates of the mileage position along the direction of the track extension. neighborhood The coordinates of the nearest location within.

7. The modular and connectable track inspection vehicle multi-device collaborative data acquisition method as described in claim 4, characterized in that, The initial conditions for the recursion are: , in, The coordinates of the mileage position along the direction of the track extension. The first chapter of the Sutra The next test uses the cumulative wear index, which is updated incrementally. The coordinates of the mileage position along the direction of the track extension. neighborhood Neighboring location coordinates First The profile deviation feature quantity that passed the test.

8. The modular and connectable track inspection vehicle multi-device collaborative data acquisition method as described in claim 1, characterized in that, The calculated comprehensive track defect risk index includes: , in, The coordinates of the mileage position along the direction of the track extension. The comprehensive defect risk index of the track. The coordinates of the mileage position along the direction of the track extension. Image damage score at the location, This serves as a normalized reference value for image damage scoring. The coordinates of the mileage position along the direction of the track extension. The cumulative wear index at the location, This serves as a normalized reference value for the cumulative wear index. The mileage coordinates extracted from the fused point cloud data along the track extension direction. The deviation between the measured track gauge and the preset standard track gauge. This serves as a normalized reference value for track gauge deviation. Image damage weights, For the weight of profile wear, As the weight for track gauge deviation, The order parameter of the weighted power average. The coordinates of the mileage position along the direction of the track extension. The cross-coupling feature between image damage score and cumulative wear index. The upper limit coefficient for saturation. It is a semi-saturated constant. It is a synergy index.

9. A modular, connectable track inspection vehicle multi-device collaborative data acquisition method as described in claim 8, characterized in that, Calculate the mileage coordinates along the track extension direction. Cross-coupling features of image damage score and cumulative wear index include: 。 10. A modular, connectable track inspection vehicle multi-device collaborative data acquisition system, the system being applied to a track inspection vehicle, the track inspection vehicle comprising a modular body, the modular body being equipped with an image acquisition module, a profile point cloud acquisition module, a positioning and attitude determination module, a mileage acquisition module, and a synchronization control module, characterized in that, include: The timing module is used by the synchronization control module to complete the timing according to the external timing signal output by the positioning and attitude determination module, and to generate a trigger signal according to the pulse signal output by the mileage acquisition module, and distribute the trigger signal to the image acquisition module and the profile point cloud acquisition module to obtain the timing trigger signal; The acquisition module is used for the image acquisition module and the contour point cloud acquisition module to respectively respond to the timed trigger signal to acquire data, obtain raw image data, raw point cloud data and their respective original timestamps, and feed the original timestamps back to the synchronization control module; The synchronization compensation module is used by the synchronization control module to dynamically compensate the acquisition time of the image acquisition module and the profile point cloud acquisition module based on the original timestamp, the running speed information output by the mileage acquisition module and the attitude angular velocity information output by the positioning and attitude determination module, so as to obtain the synchronization compensation timestamp. The fusion module is used by the synchronization control module to perform weighted fusion of the original image data and the original point cloud data based on the synchronization compensation timestamp to obtain fused image data and fused point cloud data. The cumulative wear index calculation module is used by the synchronization control module to calculate the cumulative wear index of the track profile based on the fused point cloud data. The track comprehensive defect risk index calculation module is used by the synchronization control module to calculate the track comprehensive defect risk index by combining the image damage score obtained from the fused image data with the cumulative wear index and the track gauge deviation extracted from the fused point cloud data. The maintenance decision information output module is used by the synchronous control module to compare the track comprehensive defect risk index with a preset risk threshold and output maintenance decision information.