Track comprehensive detection method based on direction mileage coordinate system

Abstract

The invention relates to a comprehensive track detection method based on a directional mileage coordinate system, which comprises the following steps of: acquiring various state data of a track in real time, and preprocessing; based on the preprocessed various state data of the track, a Unity engine is used for building a simulation scene of a simulation track; in the simulation scene, performing coordinate transformation and fusion on the preprocessed various state data of the track, and dynamically generating a directional mileage coordinate system model based on an inertial reference; and on the basis of the direction mileage coordinate system model based on the inertial reference, identifying abnormal data by adopting a fault information calibration method to obtain fault information of the track. Compared with the prior art, the method has the advantages of improving the track detection efficiency, enhancing the risk early warning capability and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of track detection, and in particular to a comprehensive track detection method based on a direction mileage coordinate system. Background Art

[0002] As an important infrastructure of the economy, the safety and comfort of rail transit are directly affected by the track status. With the rapid development of rail transit, it is particularly important to ensure the efficiency and accuracy of track safety detection. The daily maintenance of rail transit systems requires regular inspection of track status to identify potential faults and ensure safe operation. However, traditional track detection methods mainly rely on manual detection and analysis of single sensor data. These methods have problems such as low efficiency and difficulty in reflecting track status in real time. In addition, there is a lack of visualization of track status, which makes fault diagnosis and early warning difficult and cannot effectively assist maintenance personnel in making quick decisions.

[0003] Although various track inspection equipment, such as track inspection vehicles and track inspection robots, have emerged in recent years, the existing inspection methods and results display still have certain limitations: (1) Data silos: The data formats and coordinate systems collected by different sensors are not unified, making it difficult to effectively integrate and analyze them. For example, encoders, inertial navigation systems (speedometers and gyroscopes), lidars, cameras, etc., the data measured by these sensors are in different forms and are difficult to integrate. Coordinate conversion and error compensation are required to achieve more accurate data integration. (2) Difficulty in data processing: Existing track inspection methods often require a complex coordinate system construction process, which needs to consider factors such as the curvature, inclination, superelevation, and superelevation slope of the track, resulting in difficulty in coordinate system construction and low efficiency. (3) Poor visualization: There is a lack of effective visualization tools, making it difficult to intuitively and real-timely display the comprehensive status information of the track. Existing track inspection methods can often only provide two-dimensional track status display, which cannot intuitively reflect the three-dimensional structure and spatial relationship of the track, and it is difficult to meet the refined needs of track inspection. (4) Low analysis efficiency: The data analysis method is relatively simple, making it difficult to deeply explore the value of the data.

[0004] In recent years, multi-sensor detection has been widely used in various fields. However, due to the different types of collected data and real-time issues, data analysis and fusion are difficult and the display of data synthesis results is poor. The detection methods are inefficient and it is difficult to display the track status in real time. There is an urgent need to make new progress in data processing and data visualization to solve the above problems. Summary of the invention

[0005] The purpose of the present invention is to provide a track comprehensive detection method based on a direction mileage coordinate system to improve the accuracy and efficiency of track detection.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A track comprehensive detection method based on a direction mileage coordinate system comprises the following steps:

[0008] Collect various track status data in real time and perform preprocessing;

[0009] Based on the pre-processed track state data, the Unity engine is used to build a simulation scene of the simulated track;

[0010] In the simulation scenario, coordinate transformation and fusion are performed on the preprocessed track multiple state data to dynamically generate a direction mileage coordinate system model based on an inertial reference;

[0011] Based on the inertial reference-based directional mileage coordinate system model, a fault information calibration method is used to identify abnormal data to obtain track fault information.

[0012] Furthermore, the preprocessing step includes filtering and denoising processing.

[0013] Furthermore, the various status data are collected through various sensors mounted on the track inspection vehicle, including IMU, encoder, displacement sensor, lidar, and camera. The IMU measures the posture data of the track inspection vehicle when it runs on the track, the encoder measures the mileage data of the track inspection vehicle, the displacement sensor measures the geometric parameters of the track, the lidar scans and measures the point cloud data of the track space when the track inspection vehicle is running, and the camera captures the image data of the track.

[0014] Furthermore, the step of building a simulation scene of a simulated track includes:

[0015] The design parameters of the track are obtained, and combined with the geometric parameters of the preprocessed track, the Unity engine is used to build a simulation scene of the initial simulated track, and the track inspection vehicle model is added to obtain the simulation scene of the final simulated track.

[0016] Furthermore, the step of dynamically generating a direction mileage coordinate system model based on an inertial reference includes:

[0017] Integrate the preprocessed attitude data and mileage data to construct an initial dynamic directional mileage coordinate system model based on inertial reference;

[0018] The preprocessed geometric parameters, point cloud data and image data are transformed into coordinates to be mapped into the initial dynamic inertial reference-based directional mileage coordinate system model, and the Kalman filter algorithm is used for data fusion and error compensation to obtain the final dynamic inertial reference-based directional mileage coordinate system model.

[0019] Furthermore, the step of constructing an initial dynamic directional mileage coordinate system model based on an inertial reference includes:

[0020] Performing coordinate transformation on the preprocessed posture data and mileage data to obtain the coordinate transformed posture data and mileage data;

[0021] In the simulation scene of the simulated track, the position coordinates of the track inspection vehicle during operation and the accelerations in various directions at various positions are calculated based on the posture data and mileage data after the coordinate transformation, and the running trajectory is generated through the Bezier curve to form an initial dynamic direction mileage coordinate system model based on the inertial reference, wherein the arrow lengths of the x, y, and z axes in the initial dynamic direction mileage coordinate system model based on the inertial reference represent the acceleration values ​​in the three directions, respectively.

[0022] Furthermore, the coordinate transformation expression of the preprocessed posture data is:

[0023] P′=R·P+t

[0024] Where P is the original coordinate point, that is, the coordinate point in the camera coordinate system, P′ is the transformed coordinate point, that is, the coordinate point in the direction mileage coordinate system, R is the rotation matrix, which is used to represent the rotation transformation of the coordinate system, and t is the translation vector, which is used to represent the translation transformation of the coordinate system;

[0025] The coordinate transformation expression for the coordinate transformation of the pre-processed mileage data is:

[0026] s=1 / 2a·t 2

[0027] Where s is the mileage coordinate point per unit time t, and a is the acceleration in the attitude data.

[0028] Furthermore, the step of performing coordinate transformation on the preprocessed geometric parameters, point cloud data and image data comprises:

[0029] For the coordinate transformation of preprocessed geometric parameters:

[0030] The preprocessed geometric parameters are converted from the displacement sensor coordinate system to the track inspection vehicle coordinate system, where the conversion expression is:

[0031]

[0032] Where P car is the coordinate in the track inspection vehicle coordinate system, T sensor is the coordinate rotation matrix, which is used to transform from the sensor coordinate system to the track inspection vehicle coordinate system. s is the lateral offset, y s is the vertical height offset;

[0033] The geometric parameters in the track inspection vehicle coordinate system are converted to the direction mileage coordinate system, where the conversion expression is:

[0034] P dir =R car ·P car +t car

[0035] Where P dir is the coordinate point in the direction mileage coordinate system, R car ,t car Generate rotation matrix and translation vector for posture data respectively;

[0036] For the coordinate transformation of preprocessed point cloud data:

[0037] The preprocessed point cloud data is converted from the LiDAR coordinate system to the mileage coordinate system using a coordinate conversion formula, where the coordinate conversion expression is:

[0038] P′=R·P+t

[0039] Where P is the original coordinate point, that is, the coordinate point in the laser radar coordinate system, P′ is the transformed coordinate point, that is, the coordinate point in the direction mileage coordinate system, R is the rotation matrix, which is used to represent the rotation transformation of the coordinate system, and t is the translation vector, which is used to represent the translation transformation of the coordinate system;

[0040] For coordinate transformation of preprocessed image data:

[0041] The preprocessed image data is converted from the camera coordinate system to the pixel coordinate system, where the conversion expression is:

[0042] x′=f ＊ X / Z+cx,y′=f ＊ Y / Z+cy

[0043] Where x′, y′ are the coordinate points in the pixel coordinate system, X, Y and Z are the coordinate points in the camera coordinate system, f is the focal length of the camera, cx, cy are the optical centers of the camera;

[0044] The image data in the pixel coordinate system is converted to the direction mileage coordinate system, where the conversion expression is:

[0045] X=x′-cx ＊ Z / f*k1,Y=y′-cy ＊ Z / f*k2

[0046] Where k1 and k2 are radial distortion coefficients.

[0047] Furthermore, the Kalman filter algorithm is expressed as:

[0048]

[0049] K=P prior H T (HP prior H T +R) -1

[0050] In the formula, is the updated state estimate, is the predicted state estimate, K is the Kalman gain, z is the observation value, H is the observation matrix, P prior is the prediction error covariance, and R is the observation noise covariance matrix.

[0051] Furthermore, the fault information calibration method uses a fault information calibration expression to identify abnormal data in the directional mileage coordinate system model based on the inertial reference to obtain track fault information, wherein the fault information includes a fault location and a fault type, wherein the fault information calibration expression is:

[0052] F={(x,y,z)|x=x0+Δx,y=y0+Δy,z=z0+Δz}

[0053] Where F is a function expression, which represents the calibration result of the fault information; (x0, y0, z0) is the original coordinate of the fault point; (Δx, Δy, Δz) is the offset of the fault point.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] (1) The present invention integrates multiple track status data and uses the modeling function of the Unity engine to simplify the construction process, quickly construct a dynamic inertial reference-based directional mileage coordinate system model, and can provide more accurate track status information. Fault detection is performed through a fault information calibration method, which can significantly improve the accuracy and efficiency of track detection.

[0056] (2) When constructing the direction-mileage coordinate system model, the present invention first converts the data collected by all sensors into a uniform coordinate system, first constructs an initial direction-mileage coordinate system model that uses acceleration to display dynamics by integrating direction and mileage data, and fuses and compensates for errors of different sensors through the Kalman filter algorithm, maps the remaining sensor collected data to the initial direction-mileage coordinate system model, and obtains the final dynamic direction-mileage coordinate system model. The present invention can effectively fuse multiple sensor data and realize all-round visualization of track status information, which helps to improve the efficiency of track detection and the accuracy of data analysis, enhance track status identification and safety risk warning capabilities, and is of great significance for ensuring rail transit safety.

[0057] (3) The present invention can provide technical support for the field of rail transit, ensure the safe operation of tracks, reduce maintenance costs, and promote the development of rail transit towards intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0059] Figure 2 It is the conceptual principle diagram of the directional mileage coordinate system of the present invention;

[0060] Figure 3 It is a schematic diagram showing the Unity three-dimensional effect of the direction mileage coordinate system of the present invention. DETAILED DESCRIPTION

[0061] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0062] This embodiment provides a comprehensive track detection method based on a direction mileage coordinate system, such as Figure 1 As shown, the method comprises the following steps:

[0063] Step 1: Sensor data collection and processing.

[0064] By installing multiple sensors such as inertial measurement units (IMUs), encoders, lidars, 2D displacement sensors, and industrial cameras on the track inspection vehicle, data on various aspects of the track status are collected in real time. The IMU is used to measure the changes in the posture of the derailment inspection vehicle during its operation on the track. The encoder is used to measure the actual mileage of the derailment inspection vehicle, that is, the actual length of the track section it has passed. The displacement sensor measures the geometric parameters of the track. The lidar scans the track space limits during operation, and the industrial camera is used to capture the image information of the track. These data are then analyzed, filtered, and denoised to eliminate the impact of errors and ensure data accuracy and consistency. For example, the collected raw data is preliminarily processed in Python, such as Figure 2 As shown in the figure, the acceleration data is converted into position coordinate data and standardized as the input data of the directional mileage coordinate system model. After completion, the standardized data is imported into the Unity resource library for subsequent 3D display.

[0065] Step 2: Building a 3D scene.

[0066] By utilizing the 3D modeling function of the Unity engine, a simulation scene of the real track is built according to the geometric parameters and design parameters of the track, and a model of the track inspection vehicle is added to simulate the physical environment of the train operation.

[0067] Step 3: Dynamic generation of the directional mileage coordinate system model based on the inertial reference.

[0068] (1) Coordinate transformation of sensor data:

[0069] The data measured by different sensors are integrated and mapped into the mileage coordinate system through specific coordinate transformation formulas and methods to achieve the transformation between different reference systems and form a mileage coordinate system model based on inertial reference (a dynamic three-dimensional coordinate system). For example, for the point cloud data measured by LiDAR, the rotation matrix R and translation vector t are used for coordinate transformation. The coordinate transformation formula is as follows:

[0070] P′=R·P+t

[0071] Among them, P is the original coordinate point; P′ is the converted coordinate point; R is the rotation matrix, which is used to represent the rotation transformation of the coordinate system; t is the translation vector, which is used to represent the translation transformation of the coordinate system.

[0072] For the mileage data measured by the encoder, the coordinate transformation is performed, and the coordinate transformation expression is:

[0073] s=1 / 2a·t 2

[0074] Where s is the mileage coordinate point per unit time t, and a is the acceleration in the attitude data.

[0075] The specific steps are:

[0076] The attitude data is obtained according to the inertial navigation system (IMU), the rotation matrix R is constructed using Euler angles or quaternions, and the actual position of the track inspection vehicle on the track is obtained from the mileage data of the encoder to construct the translation vector t.

[0077] For the point cloud data of the lidar, the above coordinate conversion formula is used to map the point cloud data from the lidar coordinate system to the mileage coordinate system. For the image data of the industrial camera, the image feature points are corrected by the camera's internal and external parameters and then mapped to the mileage coordinate system using the coordinate conversion formula.

[0078] For the image data taken by the industrial camera, the image data taken by the camera is corrected: the internal parameter correction is to correct the distortion of the lens, and the external parameter correction is to determine the position and direction of the camera in the world coordinate system. The pixel points in the image are mapped to three-dimensional space points using the triangulation method. The mathematical model formula of the camera involved is as follows:

[0079] Conversion from camera coordinate system to pixel coordinate system: x′=f ＊ X / Z+cx,y′=f ＊ Y / Z+cy;

[0080] Conversion from pixel coordinate system to mileage coordinate system: X = x′-cx ＊ Z / f*k1,Y=y′-cy ＊ Z / f*k2

[0081] Where x′ and y′ are the coordinates in the pixel coordinate system, X, Y and Z are the coordinates in the camera coordinate system, f is the focal length of the camera, cx, cy is the optical center of the camera, and k1 and k2 are the radial distortion coefficients.

[0082] Apply the above coordinate conversion formula to convert the image feature points from the camera coordinate system to the mileage coordinate system. For the plane coordinate data of the 2D displacement sensor, to convert it to the mileage coordinate system, the main steps and formulas involved are as follows:

[0083] ① Calibrate the sensor first. Determine the installation position of the 2D displacement sensor and the posture of its coordinate system relative to the track inspection vehicle coordinate system. Record the sensor's internal parameters (measurement accuracy, error range) and installation external parameters (rotation and translation relative to the track inspection vehicle).

[0084] ②Data collection. Collect the lateral offset x from the 2D displacement sensor s and the vertical height offset y s . Filter the sensor data to remove noise caused by environmental interference.

[0085] ③ Transformation from sensor to track inspection vehicle coordinate system. Using the known T sensor Matrix, which converts the collected data from the sensor coordinate system to the track inspection vehicle coordinate system. Specific formula:

[0086]

[0087] ④Conversion from the track inspection vehicle to the mileage coordinate system. Use the track inspection vehicle attitude data provided by the IMU to generate the rotation matrix R car and the translation vector t car . Point P in the track inspection vehicle coordinate system car Mapping to the direction mileage coordinate system:

[0088] P dir =R car ·P car +t car

[0089] ⑤ Data standardization. Adjust the units of the converted data (e.g. convert millimeters to meters). Ensure that the timestamps of the data are aligned so that they can be fused with data from other sensors.

[0090] (2) Constructing a directional mileage coordinate system model based on inertial reference:

[0091] During construction, the initial direction and mileage coordinate system model based on inertial reference is first constructed by integrating the direction and mileage data after coordinate conversion, that is, the attitude data measured by the IMU and the mileage measured by the encoder. Then, the data collected by other sensors after coordinate conversion is mapped to the initial direction and mileage coordinate system model through data fusion to form the final direction and mileage coordinate system model based on inertial reference. Specifically:

[0092] After standardizing the sensor data in Python, such as using an existing Excel spreadsheet of inertial navigation data (including initial velocity, acceleration, etc.), import it into Python for programming and running, and calculate the position coordinates of each unit time of the running trajectory. Load the position data of the track inspection vehicle into the Resources folder of Unity, compile and run it through C# code in visual studio, read the data in the Resources folder and make the target object move according to the coordinates in the data, and also compile the C# code to make the camera follow the movement of the target object to follow and observe the driving status of the track inspection vehicle on the track. Attach the C# code to Unity, and attach it to the target object to make it an object script, then click Play, you can see that the car moves according to the specified trajectory and the data changes on the Y-axis are represented by the vibration of the car (moving trajectory and moving speed can be adjusted). Figure 2As shown in the figure, by integrating the measured inertial reference and mileage data, a dynamic initial direction mileage coordinate system model with the center point of the track detection vehicle as the origin of the coordinate system is constructed in Unity, so that the dynamic coordinate system moves according to the coordinates in the data. In this coordinate system, the X-axis represents the lateral distance of the track, the Y-axis represents the longitudinal distance of the track, and the Z-axis represents the height of the track. The lengths of the arrows on the X, Y, and Z axes represent the acceleration magnitudes in the three directions (a x 、a y 、a z ), reflects the state of the track by showing the different speeds, postures, vibrations and other information of the inspection vehicle under different track conditions in a dynamic coordinate system in real time. This coordinate system is used to describe the spatial position and direction information of the track, including elements such as the track centerline, mileage scale, mileage point, and origin of the coordinate system. It can reflect the curvature, inclination, superelevation, superelevation along the slope and other characteristics of the track in real time, providing a data base for the full visualization of the track status. The running track is generated through the Bezier curve to show the dynamic effect of the track inspection vehicle running along the track, and it is dynamically updated to ensure that the track status is reflected in real time during the track inspection process.

[0093] Finally, data fusion is performed: During the data fusion process, the Kalman filter algorithm is used to fuse and compensate for the errors of the data from different sensors to improve the accuracy of the data, and the data is displayed in a three-dimensional directional mileage coordinate system to obtain a directional mileage coordinate system model based on an inertial reference, thereby realizing the full-scale visualization of the track comprehensive condition detection data. The formula of the Kalman filter algorithm is as follows:

[0094]

[0095] in, is the updated state estimate; is the predicted state estimate; K is the Kalman gain.

[0096] The calculation formula of K is:

[0097] K=P prior H T (HP prior H T +R) -1

[0098] Among them, z is the observation value; H is the observation matrix; P prior is the prediction error covariance; R is the observation noise covariance matrix.

[0099] The final directional mileage coordinate system Unity 3D effect is shown as follows Figure 3 shown.

[0100] This step effectively solves the data inconsistency and error problems by using the Kalman filter algorithm and coordinate transformation formula, and enhances the reliability and accuracy of track detection.

[0101] Step 4: Fault information calibration and risk warning.

[0102] The fault information calibration formula is used to analyze the data, identify abnormal data, and obtain the fault information of the track. The characteristic position is calibrated and the early warning is displayed. The fault position and type of the track are dynamically displayed in the direction mileage coordinate system, including information such as curves, slopes, superelevation, and track unevenness, so that maintenance personnel can take timely measures. For the detected fault point, the following fault information calibration formula is used for position calibration:

[0103] F={(x,y,z)|x=x0+Δx,y=y0+Δy,z=z0+Δz}

[0104] Among them, (x0, y0, z0) is the original coordinate of the fault point; (Δx, Δy, Δz) is the offset of the fault point.

[0105] By visually warning the fault points, it is easy to find safety risks in time and ensure the safe operation of the track. The data analysis results can also provide a scientific basis for track maintenance and support decision-making. It not only improves the efficiency and accuracy of track detection, but also enhances the ability to identify track status and warn of safety risks, which is of great significance to ensuring the safety of rail transportation.

[0106] Step 5: Augmented reality interaction.

[0107] By using Unity's 2D and 3D view switching function, users can select the appropriate view to observe the comprehensive status of the track as needed, and dynamically display the track, posture and vibration of the track inspection vehicle in an all-round three-dimensional effect. For example, users can view the overall status of the track in a 2D view, and view the specific fault point and its spatial location in a 3D view. Through 3D visualization, the track status data is displayed in the form of a 3D model, and the inspection personnel can intuitively see the status of the track, quickly locate potential fault points, and improve the efficiency of fault location and analysis. It also supports users to interact with the virtual environment, such as switching perspectives, adjusting model parameters, and querying track information. The augmented reality (AR) function allows the virtual track model to be superimposed on the real scene, thereby enhancing the understanding and perception of the track status.

[0108] The addition of augmented reality interactive functions can intuitively display the three-dimensional structure and spatial relationship of the track, making the display of the track status more vivid and intuitive, and providing effective technical support for inspection and maintenance.

[0109] In summary, this embodiment provides a method for constructing a direction-mileage coordinate system for comprehensive track detection based on the Unity engine, which mainly includes: collection and processing of sensor data, construction of three-dimensional scenes, dynamic generation of direction-mileage coordinate system models, coordinate conversion and fusion of sensor data, fault information calibration and risk warning, and augmented reality interaction; collecting a variety of data by installing different sensors on the track inspection vehicle, and analyzing and processing the data; building a simulation scene simulating the real track in Unity; constructing a dynamic direction-mileage coordinate system model by integrating direction and mileage data; integrating different data and displaying them in the equation-mileage coordinate system model through specific coordinate conversion formulas; analyzing the data to obtain fault information and calibrating and warning display; and realizing the interaction of augmented reality scenes by switching between Unity's two-dimensional and three-dimensional perspectives.

[0110] The method of the present invention can effectively integrate multiple sensor data to achieve all-round visualization of track status information. This method not only improves the efficiency of track detection and the accuracy of data analysis, but also enhances the track status identification and safety risk warning capabilities, which is of great significance for ensuring the safety of rail transit. The practical application value of the present invention lies in that it can provide technical support for the field of rail transit, ensure the safe operation of the track, reduce maintenance costs, and promote the development of rail transit towards intelligence. In general, the present invention has significant advantages and broad application prospects in improving the maintenance efficiency and safety of rail transit systems.

[0111] If the above functions are implemented in the form of software functional units and sold or used as independent products, they 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 the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0112] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0113] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0114] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0115] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0116] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0117] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A track comprehensive detection method based on a direction mileage coordinate system, characterized in that: The following steps are involved: Collect various track status data in real time and perform preprocessing; Based on the pre-processed track state data, the Unity engine is used to build a simulation scene of the simulated track; In the simulation scenario, coordinate transformation and fusion are performed on the preprocessed track multiple state data to dynamically generate a direction mileage coordinate system model based on an inertial reference; Based on the inertial reference-based directional mileage coordinate system model, a fault information calibration method is used to identify abnormal data to obtain track fault information.

2. A track comprehensive detection method based on a direction mileage coordinate system according to claim 1, characterized in that: The pre-processing step includes filtering and denoising.

3. A track comprehensive detection method based on a direction mileage coordinate system according to claim 1, characterized in that: The various status data are collected through various sensors carried on the track inspection vehicle, including IMU, encoder, displacement sensor, laser radar, and camera. The IMU measures the posture data of the track inspection vehicle when it runs on the track, the encoder measures the mileage data of the track inspection vehicle, the displacement sensor measures the geometric parameters of the track, the laser radar scans and measures the point cloud data of the track space when the track inspection vehicle is running, and the camera captures the image data of the track.

4. A track comprehensive detection method based on a direction mileage coordinate system according to claim 3, characterized in that: The steps of constructing a simulation scene of a simulated track include: The design parameters of the track are obtained, and combined with the geometric parameters of the preprocessed track, the Unity engine is used to build a simulation scene of the initial simulated track, and the track inspection vehicle model is added to obtain the simulation scene of the final simulated track.

5. The track comprehensive detection method based on the direction mileage coordinate system according to claim 3 is characterized in that: The step of dynamically generating a direction mileage coordinate system model based on an inertial reference comprises: Integrate the preprocessed attitude data and mileage data to construct an initial dynamic directional mileage coordinate system model based on inertial reference; The preprocessed geometric parameters, point cloud data and image data are transformed into coordinates to be mapped into the initial dynamic inertial reference-based directional mileage coordinate system model, and the Kalman filter algorithm is used for data fusion and error compensation to obtain the final dynamic inertial reference-based directional mileage coordinate system model.

6. A track comprehensive detection method based on a direction mileage coordinate system according to claim 5, characterized in that: The step of constructing an initial dynamic directional mileage coordinate system model based on an inertial reference comprises: Performing coordinate transformation on the preprocessed posture data and mileage data to obtain the coordinate transformed posture data and mileage data; In the simulation scene of the simulated track, the position coordinates of the track inspection vehicle during operation and the accelerations in various directions at various positions are calculated based on the posture data and mileage data after the coordinate transformation, and the running trajectory is generated through the Bezier curve to form an initial dynamic direction mileage coordinate system model based on the inertial reference, wherein the arrow lengths of the x, y, and z axes in the initial dynamic direction mileage coordinate system model based on the inertial reference represent the acceleration values ​​in the three directions, respectively.

7. A track comprehensive detection method based on a direction mileage coordinate system according to claim 6, characterized in that: The coordinate transformation expression of the preprocessed posture data is: P′=R·P+t Where P is the original coordinate point, that is, the coordinate point in the camera coordinate system, P′ is the transformed coordinate point, that is, the coordinate point in the direction mileage coordinate system, r is the rotation matrix, which is used to represent the rotation transformation of the coordinate system, and t is the translation vector, which is used to represent the translation transformation of the coordinate system; The coordinate transformation expression for the coordinate transformation of the pre-processed mileage data is: s=1 / 2a·t 2 Where s is the mileage coordinate point per unit time t, and a is the acceleration in the attitude data.

8. The track comprehensive detection method based on the direction mileage coordinate system according to claim 5 is characterized in that: The step of performing coordinate transformation on the preprocessed geometric parameters, point cloud data and image data comprises: For the coordinate transformation of preprocessed geometric parameters: The preprocessed geometric parameters are converted from the displacement sensor coordinate system to the track inspection vehicle coordinate system, where the conversion expression is: Where P car is the coordinate in the track inspection vehicle coordinate system, T sensor is the coordinate rotation matrix, which is used to transform from the sensor coordinate system to the track inspection vehicle coordinate system. s is the lateral offset, y s is the vertical height offset; The geometric parameters in the track inspection vehicle coordinate system are converted to the direction mileage coordinate system, where the conversion expression is: P dir =R car ·P car +t car Where P dir is the coordinate point in the direction mileage coordinate system, R car ,t car Generate rotation matrix and translation vector for posture data respectively; For the coordinate transformation of preprocessed point cloud data: The preprocessed point cloud data is converted from the LiDAR coordinate system to the mileage coordinate system using a coordinate conversion formula, where the coordinate conversion expression is: P′=R·P+t Where P is the original coordinate point, that is, the coordinate point in the laser radar coordinate system, P′ is the transformed coordinate point, that is, the coordinate point in the direction mileage coordinate system, R is the rotation matrix, which is used to represent the rotation transformation of the coordinate system, and t is the translation vector, which is used to represent the translation transformation of the coordinate system; For coordinate transformation of preprocessed image data: The preprocessed image data is converted from the camera coordinate system to the pixel coordinate system, where the conversion expression is: x′=f ＊ X / Z+cx,y′=f ＊ Y / Z+cy Where x′, y′ are the coordinate points in the pixel coordinate system, X, Y and Z are the coordinate points in the camera coordinate system, f is the focal length of the camera, cx, cy are the optical centers of the camera; The image data in the pixel coordinate system is converted to the direction mileage coordinate system, where the conversion expression is: X=x′-cx ＊ Z / f*k1,Y=y′-cy ＊ z / f*k2 Where k1 and k2 are radial distortion coefficients.

9. The track comprehensive detection method based on the direction mileage coordinate system according to claim 5 is characterized in that: The Kalman filter algorithm is expressed as: In the formula, is the updated state estimate, is the predicted state estimate, K is the Kalman gain, z is the observation value, H is the observation matrix, P prior is the prediction error covariance, and R is the observation noise covariance matrix.

10. The track comprehensive detection method based on the direction mileage coordinate system according to claim 1 is characterized in that: The fault information calibration method uses a fault information calibration expression to identify abnormal data in the directional mileage coordinate system model based on the inertial reference to obtain track fault information, wherein the fault information includes a fault location and a fault type, wherein the fault information calibration expression is: F={(x,y,z)|x=x0+Δx,y=y0+Δy,z=z0+Δz} Where F is a function expression, which represents the calibration result of the fault information; (x0, y0, z0) is the original coordinate of the fault point; (Δx, Δy, Δz) is the offset of the fault point.

Images