Vibration compensation method and device, electronic equipment, train, medium and product

Abstract

The present disclosure provides a vibration compensation method, device, electronic equipment, train, medium and product. Relates to the field of rail transit, the above-mentioned method comprises: according to the pre-constructed pose change relationship of the pose of the car body changing with time under the influence of vibration, the vibration pose of the car body generated by the car body under the influence of vibration at the to-be-measured time point is determined, the pose change relationship is constructed by using a plurality of first offset data of the relative position between the contact rail and the running rail of the train changing with time and a plurality of second offset data of the relative position between the car body and the running rail changing with time; based on the car body vibration pose, the motion displacement of the car body relative to the running rail at the to-be-measured time point is determined; the motion displacement of the car body is converted into the car body vibration compensation parameter; the train is compensated based on the car body vibration compensation parameter. The application can more accurately compensate the train vibration.

Description

Technical Field

[0001] This application relates to the field of rail transit technology, and more specifically, to a vibration compensation method, device, electronic equipment, train, medium, and product. Background Technology

[0002] In rail transit, the reliability and safety of trains are becoming increasingly important. Trains generate vibrations during operation, which not only adversely affect the train itself but also the running rails and contact rails, thus impacting the train's reliability and safety. Therefore, vibration compensation is necessary. However, the accuracy of vibration compensation using existing technologies is relatively low. Summary of the Invention

[0003] In view of this, this application provides a vibration compensation method, apparatus, electronic device, train, medium, and product.

[0004] One aspect of this application provides a vibration compensation method, comprising: determining the vehicle body vibration pose at a test time point based on a pre-constructed pose change relationship of the vehicle body pose under vibration influence over time, wherein the pose change relationship is constructed based on multiple sets of first offset data and multiple sets of second offset data acquired by a data acquisition device installed on the vehicle body during train operation, wherein the multiple sets of first offset data and multiple sets of second offset data are respectively: multiple sets of offset data on the relative position between the contact rail and the running rail of the train changing over time under the influence of vehicle body vibration and the influence of vehicle body vibration on the data acquisition device, and multiple sets of offset data on the relative position between the vehicle body and the running rail changing over time; determining the motion displacement of the vehicle body relative to the running rail at the test time point based on the vehicle body vibration pose; converting the motion displacement into vehicle body vibration compensation parameters; and performing vibration compensation on the train based on the vehicle body vibration compensation parameters.

[0005] According to an embodiment of this application, the data acquisition device includes: a first laser, a first camera, a second laser, and a second camera; wherein, the first offset data is acquired in the following manner: using the first camera to acquire a first image of the contact rail irradiated by the first laser; based on the line laser on the contact rail in the first image, determining the first offset data of the contact rail relative to the running rail in a cross section perpendicular to the track extension direction in the world coordinate system; wherein, the second offset data is acquired in the following manner: using the second camera to acquire multiple second images of the running rail irradiated by the second laser; based on the pixel coordinates of the line laser on the running rail in the multiple second images, determining the laser position of the line laser corresponding to the multiple second images in the world coordinate system; and determining the second offset data of the relative position between the vehicle body and the running rail based on the laser positions corresponding to the multiple second images.

[0006] According to an embodiment of this application, determining the first offset data of the contact rail relative to the running rail in a cross section perpendicular to the track extension direction in the world coordinate system based on the line laser on the contact rail in the first image includes: performing distortion correction processing on the first image to obtain a corrected image; using the line laser on the contact rail in the first image to determine the key point pixel coordinates of key points on the contact rail in the corrected image; determining the three-dimensional spatial offset between the standard positions of the contact rail and the contact rail in the camera coordinate system based on the key point pixel coordinates; and determining the first offset data of the contact rail relative to the running rail in a cross section perpendicular to the track extension direction in the world coordinate system based on the three-dimensional spatial offset, the geometric parameters of the contact rail, and the depth information of the corrected image.

[0007] According to an embodiment of this application, a distortion correction process is performed on a first image to obtain a corrected image, including: preprocessing the first image to obtain a preprocessed image; extracting edges from the preprocessed image to obtain the edge box of the laser on the contact rail in the preprocessed image; performing N distortion correction processes on the preprocessed image to obtain the corrected image, where N is an integer greater than or equal to 1, and the nth distortion correction process includes: deforming the preprocessed image based on the pixel coordinates of the vertices of the nth edge box to obtain a deformed image, where n = 1, 2, ..., N; determining the rectangularity of the edge box in the deformed image, where the rectangularity characterizes the degree to which the shape of the edge box is close to a rectangle; if the rectangularity characterizes that the edge box in the deformed image meets the rectangle requirement, the nth distortion correction process is completed, and the deformed image is determined as the corrected image; if the rectangularity characterizes that the edge box in the deformed image does not meet the rectangle requirement, the nth edge box in the deformed image is refitted to obtain the (n+1)th edge box, and the (n+1)th distortion correction process is performed.

[0008] According to an embodiment of this application, a deformed image is obtained by deforming a preprocessed image based on the pixel coordinates of the vertices of the nth bounding box, including: determining the tilt feature of the camera when capturing the first image based on the pixel coordinates of the vertices of the nth bounding box; performing a first deformation correction on the preprocessed image based on the tilt feature to obtain a preliminary corrected image; and performing a second deformation correction on the preliminary corrected image based on the distortion feature of the camera that captured the first image and the camera intrinsic parameters to achieve deformation removal and obtain the deformed image.

[0009] According to an embodiment of this application, preprocessing a first image to obtain a preprocessed image includes: downsampling and filtering the first image to obtain a lower layer image; interpolating and dilating the lower layer image to obtain a dilated image; determining a multi-detail layer image based on the difference in pixel values ​​between the lower layer image and the dilated image; fusing the multi-detail layer images to retain detail information in each detail layer to obtain a detail fusion result; reconstructing the image based on the detail fusion result and the multi-detail layer images to obtain a reconstructed image; performing adaptive illumination compensation processing on the reconstructed image to obtain an illumination-compensated image; and marking connected regions in the illumination-compensated image to obtain a preliminary processed image.

[0010] According to an embodiment of this application, adaptive illumination compensation processing is performed on a reconstructed image to obtain an illumination-compensated image, including: extracting the background from the reconstructed image to obtain a target object image after removing the background from the reconstructed image; performing M illumination compensation processes, where M is an integer greater than or equal to 1, and the m-th illumination compensation process includes: determining a first non-uniformity coefficient characterizing the degree of gray-level non-uniformity in the target object image; performing illumination compensation on the target object image based on the m-th illumination compensation coefficient characterizing the intensity of illumination supplementation to obtain a preliminary illumination-compensated image, where m = 1, 2, ..., M; determining a second non-uniformity coefficient characterizing the degree of gray-level non-uniformity in the preliminary illumination-compensated image; if the preliminary illumination-compensated image meets the illumination compensation requirements based on the first and second non-uniformity coefficients, then the m-th illumination compensation process is completed, and the preliminary illumination-compensated image is determined as the illumination-compensated image; if the preliminary illumination-compensated image does not meet the illumination compensation requirements based on the first and second non-uniformity coefficients, then the (m+1)-th illumination compensation coefficient is determined, and the (m+1)-th compensation process is performed.

[0011] According to an embodiment of this application, the key point pixel coordinates of key points on the contact rail in the calibration image are determined using a laser line on the contact rail in the first image. This includes: extracting the contact rail region in the calibration image to obtain a contact rail image; reconstructing a three-dimensional model of the contact rail based on the pixel coordinates of the laser line in the contact rail image to obtain a three-dimensional model of the contact rail in the world coordinate system; and determining the sub-pixel level contact rail pixel coordinates of the key points of the contact rail in the calibration image based on the three-dimensional coordinates of the contact rail key points where the calibration area intersects with the laser line in the three-dimensional model of the contact rail, the correspondence between the pre-calibrated three-dimensional coordinates and the sub-pixel coordinates.

[0012] According to an embodiment of this application, the method further includes: determining the measurement error caused by vehicle vibration on the detection of the position parameters of the contact rail based on the first offset number, the second offset data, and the error law characterizing the actual position of the contact rail and the observed position of the contact rail, wherein the actual position of the contact rail is the position of the contact rail when only affected by vehicle vibration, and the observed position of the contact rail is the position of the contact rail acquired by the data acquisition device when both the vehicle body and the data acquisition device are affected by vehicle vibration; determining multiple sets of vehicle body vibration poses caused by vibration when the vibration causes measurement errors in the data acquisition device, based on multiple sets of first offset data and the measurement errors corresponding to multiple sets of first offset data; and constructing a pose change relationship based on the multiple sets of vehicle body vibration poses.

[0013] According to an embodiment of this application, based on multiple sets of first offset data and the measurement errors corresponding to the multiple sets of first offset data, determining multiple sets of vehicle body vibration poses generated by the vehicle body under the condition of measurement error includes: modeling the multiple sets of first offset data and the measurement errors corresponding to the multiple sets of first offset data as random variables, wherein the random variables represent the uncertainty of the first offset data collected under the condition of vehicle body being affected by vibration and data acquisition equipment being affected by vehicle body vibration; obtaining the pose form representation of the random variables corresponding to the multiple sets of first offset data as the multiple sets of vehicle body vibration poses generated by the vehicle body under the condition of measurement error.

[0014] According to an embodiment of this application, determining the motion displacement of the vehicle body relative to the running track at the time point to be measured based on the vehicle body vibration pose includes: performing bandpass filtering on the vehicle body vibration pose to obtain the low-frequency and high-frequency components of the vibration; converting the desired pose into an equivalent excitation quantity characterizing the vibration intensity experienced by the vehicle body based on the low-frequency and high-frequency components of the vibration; and determining the motion displacement of the vehicle body relative to the running track at the time point to be measured corresponding to the equivalent excitation quantity according to the pre-constructed force-displacement relationship between the equivalent excitation quantity and the motion displacement, wherein the force-displacement relationship is constructed based on multiple historically acquired motion displacements and multiple equivalent excitation quantities corresponding to the multiple motion displacements.

[0015] According to an embodiment of this application, a force-displacement relationship is constructed based on multiple historically acquired motion displacements and multiple equivalent excitation quantities corresponding to these motion displacements in the following manner: Based on the multiple motion displacements and their corresponding equivalent excitation quantities, the correlation between the equivalent excitation quantities and the motion displacements is extracted; based on the correlation, the influence weight of the equivalent excitation quantity on the motion displacement is determined; Z-fold relationship construction processing is performed, where Z is an integer greater than or equal to 1. The z-th relationship construction processing includes: based on the equivalent excitation quantity, the correlation, and the z-th influence weight, constructing candidate correspondences with the equivalent excitation quantity as the independent variable and the motion displacement as the dependent variable, where z = 1, 2, ..., Z; determining candidate motion displacements at test time points based on the candidate correspondences; if the prediction error rate of the candidate motion displacements determined based on the pre-built error prediction model is higher than a preset error threshold, then the (z+1)-th adjustment influence weight is determined based on the correlation, and the (z+1)-th relationship construction processing is performed; if the prediction error rate is less than or equal to the preset error threshold, then the z-th relationship construction processing is completed, and the candidate correspondences are determined as force-displacement relationships.

[0016] According to an embodiment of this application, converting motion displacement into vehicle vibration compensation parameters includes: establishing a mapping relationship between the force and displacement transmission characteristics of the vehicle active suspension system to construct a parameter conversion relationship between motion displacement and vehicle vibration compensation parameters; and converting motion displacement into vehicle vibration compensation parameters based on the parameter conversion relationship.

[0017] Another aspect of this application provides a vibration compensation device, comprising: a vehicle body vibration pose determination module, used to determine the vehicle body vibration pose at the time point to be measured by vibration based on a pre-constructed pose change relationship of the vehicle body pose under vibration over time, wherein the pose change relationship is constructed based on multiple sets of first offset data and multiple sets of second offset data collected by a data acquisition device installed on the vehicle body during train operation, wherein the multiple sets of first offset data and multiple sets of second offset data are respectively: multiple sets of offset data on the relative position between the contact rail and the running rail of the train changing over time under the influence of vehicle body vibration and the influence of vehicle body vibration on the data acquisition device, and multiple sets of offset data on the relative position between the vehicle body and the running rail changing over time; a motion displacement determination module, used to determine the motion displacement of the vehicle body relative to the running rail at the time point to be measured based on the vehicle body vibration pose; a compensation parameter conversion module, used to convert the motion displacement into vehicle body vibration compensation parameters; and a vibration compensation module, used to perform vibration compensation on the train based on the vehicle body vibration compensation parameters.

[0018] Another aspect of this application provides an electronic device comprising:

[0019] One or more processors;

[0020] Memory, used to store one or more programs.

[0021] Specifically, when one or more programs are executed by one or more processors, the one or more processors implement the above method.

[0022] Another aspect of this application provides a train, including the aforementioned electronic equipment.

[0023] Another aspect of this application provides a computer-readable storage medium storing computer-executable instructions that, when executed, are used to implement the method described above.

[0024] Another aspect of this application provides a computer program product including computer-executable instructions that, when executed, are used to implement the methods described above.

[0025] As can be seen from the above, in the solution provided in this application, by utilizing multiple sets of first offset data on the relative position between the train's contact rail and running rail changing over time, it is possible to analyze the impact of time-varying vehicle body vibration and movement on the position and geometry of the contact rail. By utilizing multiple sets of second offset data on the relative position between the vehicle body and running rail changing over time, it is possible to analyze the impact of time-varying vehicle body vibration and movement on vehicle body offset and data acquisition equipment. Combining the first and second offset data, it is possible to deduce the vehicle body pose, excluding the data acquisition equipment affected by vehicle body vibration, and only affected by vehicle body vibration. This established pose change relationship can reduce the acquisition error caused by the vehicle body vibration affecting the data acquisition equipment. Through this pose change relationship, it is also possible to predict the vehicle body vibration pose at the time to be measured in a timely manner, thereby accurately determining the vehicle body's motion displacement. This allows for real-time and more accurate determination of the vehicle body's motion displacement at the time to be measured. Using the vehicle body vibration compensation parameters obtained from the motion displacement, vibration compensation for the train can be performed more accurately. Attached Figure Description

[0026] The above and other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0027] Figure 1 The illustrations depict exemplary system architectures to which the vibration compensation methods, apparatuses, electronic devices, trains, media, and products of this application can be applied;

[0028] Figure 2 The schematic diagram illustrates a flow chart of a vibration compensation method according to an embodiment of this application;

[0029] Figure 3 The schematic diagram illustrates a flowchart of a method for determining a first offset parameter according to an embodiment of this application;

[0030] Figure 4 The schematic diagram illustrates a flow chart of an illumination compensation processing method according to an embodiment of this application;

[0031] Figure 5 The schematic diagram illustrates a process flow of a distortion correction method according to an embodiment of this application;

[0032] Figure 6 The illustration shows a flowchart of a pose change relationship calculation method according to an embodiment of this application;

[0033] Figure 7 The illustration shows a flowchart of a method for constructing a force-displacement relationship according to an embodiment of this application;

[0034] Figure 8 The schematic diagram illustrates a structural schematic of a vibration compensation device according to an embodiment of this application;

[0035] Figure 9 A block diagram of an electronic device suitable for vibration compensation methods provided in an embodiment of this application is illustrated schematically. Detailed Implementation

[0036] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0037] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0038] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0039] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0040] In the embodiments of this application, the collection, updating, analysis, processing, use, transmission, provision, disclosure, and storage of data (e.g., including but not limited to user personal information) comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures have been taken to prevent unauthorized access to user personal information data and to safeguard user personal information security, network security, and national security.

[0041] In the embodiments of this application, the user's authorization or consent was obtained before obtaining or collecting the user's personal information.

[0042] The execution subject of the embodiments of this application will be described below.

[0043] The solutions provided in this application can be applied to electronic devices such as desktop computers, laptops, tablets, and electronic control units. For ease of description, the subject matter of the vibration compensation method provided in this application is collectively referred to as vibration compensation device.

[0044] Scenario 1

[0045] See Figure 1 An exemplary system architecture 100 is provided, which can apply the vibration compensation methods, apparatus, electronic devices, trains, media, and products provided in the embodiments of this application. For example... Figure 1 As shown, the system architecture 100 of this application embodiment may include a train 101, a network 102, and a vibration compensation device 103. The network 102 serves as a medium for providing a communication link between the train 101 and the vibration compensation device 103. The network 102 may include various connection types, such as wired and / or wireless communication links, etc. The vibration compensation device 103 can acquire multiple sets of first offset data and multiple sets of second offset data through the network 102, and send vehicle vibration compensation parameters to the train 101 through the network 102 to control the actuators installed on the train 101 to perform vibration compensation on the train.

[0046] It is important to note that Figure 1 The examples shown are merely examples of system architectures that can be applied to the embodiments of this application, in order to help those skilled in the art understand the technical content of this application, but do not mean that the embodiments of this application cannot be used in other devices, systems, environments or scenarios.

[0047] Scenario 2

[0048] Vibration compensation equipment can be installed on the train. It can communicate with data acquisition equipment and actuators installed on the train to acquire first and second offset data collected by the data acquisition equipment, and control the actuators to perform vibration compensation on the train based on the vehicle body vibration compensation parameters.

[0049] The vibration compensation method of the present application embodiment will be described in detail below.

[0050] According to an embodiment of this application, see Figure 2 A flowchart of a vibration compensation method is provided, which includes the following operations S201-S204.

[0051] In operation S201: Based on the pre-constructed relationship between the vehicle body pose under vibration and time, determine the vehicle body vibration pose generated by the vibration at the time point to be measured.

[0052] Among them, the position change relationship is constructed based on multiple sets of first offset data and multiple sets of second offset data collected by the data acquisition equipment installed on the train body under the train's running state. The multiple sets of first offset data are: multiple sets of offset data on the relative position between the train's contact rail and running rail changing over time under the influence of vibration on the train body and the influence of vibration on the data acquisition equipment.

[0053] The first data acquisition device used to collect multiple sets of first offset data and the second data acquisition device used to collect multiple sets of second offset data can be different. The first data acquisition device can be installed on the train body, for example, at least one of the following locations: the undercarriage of the train, the side of the train body, or the roof of the train.

[0054] The first offset data may include the relative position between the contact rail and the travel rail's track plane. The track plane is the plane containing the top surface of the travel rail.

[0055] For example, the first offset data may include: the deviation of the contact rail position from a preset standard contact rail position in the transverse direction of the track, wherein the transverse direction of the track is the direction of the intersection between the cross section perpendicular to the track extension direction and the track plane on the running rail. The first offset data may also include: the vertical distance between the contact rail position and the track plane. The track extension direction may be a direction parallel to the train's travel direction on the local track, determined based on a local track including the running rail and / or the contact rail. The vertical distance between the contact rail position and the track plane is: the length of the perpendicular line between the contact rail and the track plane in the aforementioned cross section.

[0056] In one implementation, the first data acquisition device may include a vibration sensor. For example, it may include a vibration acceleration sensor. The first data acquisition device acquires first offset data at a preset acquisition frequency and records the acquisition timestamps of the first offset data acquired at each time point. The time point at which the first offset data was acquired is determined based on the acquisition timestamps. The first offset data acquired at the same time point is defined as a set of first offset data. A set of first offset data may include multiple sets of first offset data or a single set of first offset data. The vibration compensation device can acquire multiple sets of first offset data acquired by the first data acquisition device.

[0057] The second offset data is collected by data acquisition equipment installed on the train body while the train is in motion. Multiple sets of second offset data are: multiple sets of offset data on the relative position between the car body and the running rail changing over time under the influence of car body vibration and the influence of car body vibration on the data acquisition equipment.

[0058] The second data acquisition device used to collect multiple sets of second offset data can be one or more. The second data acquisition device can be installed on the train body, for example, at least one location including the train undercarriage, the side of the train body, and the roof of the train. The second data acquisition device can also be installed on the contact rail.

[0059] The second offset data may include the deviation between the position of the vehicle body and the running track in the lateral direction of the track.

[0060] The acquisition method for the first offset data is similar to that for the second offset data. The difference lies in the first offset data and the second offset data, as well as the first data acquisition device and the second data acquisition device, which will not be detailed here.

[0061] The method for determining the vibration posture of the vehicle body is explained below.

[0062] Given the established pose change relationship, inputting the time point to be measured into the aforementioned pose change relationship yields the vehicle vibration pose.

[0063] Furthermore, the pose change relationship can include the distribution of the vehicle body pose caused by vibration, which is constructed based on the time-conditional Gaussian mixture model. The vibration compensation device can determine a time window of a preset time length according to the time point to be measured, and determine the pose corresponding to each time point in the above time window according to the pose change relationship. The average of the poses corresponding to each time point is taken as the vehicle body vibration pose.

[0064] The pose change relationship theoretically excludes the influence of vehicle vibration on the data acquisition equipment. Theoretically, the pose change relationship represents the pose of the vehicle body at different time points caused only by vibration. However, due to errors in actual calculations, it is difficult to completely exclude the influence of vehicle vibration on the data acquisition equipment.

[0065] In operation S202: Based on the vehicle body vibration posture, determine the motion displacement of the vehicle body relative to the running track at the time point to be measured.

[0066] The relative position between the vehicle body and the running track can include the direction of the motion displacement. For example, the direction of the motion displacement can include: the vehicle body shifting to the left or to the right in the lateral direction of the track.

[0067] In one implementation, a grey prediction model is constructed with the vehicle body vibration pose as the independent variable and the motion displacement between the vehicle body and the running track as the dependent variable. The motion displacement of the vehicle body relative to the running track at the time point corresponding to the vehicle body vibration pose is obtained through the constructed grey prediction model.

[0068] In addition, during train movement, multiple sets of first offset data and multiple sets of second offset data can be continuously collected in operation S201. The pose change relationship is periodically reconstructed according to the preset pose change relationship update frequency. When performing vibration compensation during train movement, the newly constructed pose change relationship can be used for vibration compensation.

[0069] In operation S203: convert the motion displacement into vehicle vibration compensation parameters.

[0070] In one implementation, the vibration compensation device converts the direction of the motion displacement to obtain the motion displacement in the opposite direction, and determines the motion displacement in the opposite direction as the vehicle body vibration compensation parameter.

[0071] In operation S204: Vibration compensation of the train is performed based on the vehicle body vibration compensation parameters.

[0072] In one implementation, the vibration compensation device can send vehicle body vibration compensation parameters to the actuator performing vibration compensation to control the actuator to perform vibration compensation on the train. The actuator may include at least one of the following: an active suspension actuator, an electromechanical actuator, a piezoelectric stack actuator, and a centrifugal actuator.

[0073] As can be seen from the above, in the solution provided in this application, by utilizing multiple sets of first offset data on the relative position between the train's contact rail and running rail changing over time, it is possible to analyze the impact of time-varying vehicle body vibration and movement on the position and geometry of the contact rail. By utilizing multiple sets of second offset data on the relative position between the vehicle body and running rail changing over time, it is possible to analyze the impact of time-varying vehicle body vibration and movement on vehicle body offset and data acquisition equipment. Combining the first and second offset data, it is possible to deduce the vehicle body pose, excluding the data acquisition equipment affected by vehicle body vibration, and only affected by vehicle body vibration. This established pose change relationship can reduce the acquisition error caused by the vehicle body vibration affecting the data acquisition equipment. Through this pose change relationship, it is also possible to predict the vehicle body vibration pose at the time to be measured in a timely manner, thereby accurately determining the vehicle body's motion displacement. This allows for real-time and more accurate determination of the vehicle body's motion displacement at the time to be measured. Using the vehicle body vibration compensation parameters obtained from the motion displacement, it is possible to more accurately compensate for train vibration.

[0074] The following sections will explain another method for collecting multiple sets of first offset data and multiple sets of second offset data.

[0075] In one embodiment of this application, the data acquisition device includes: a first laser, a first camera, a second laser, and a second camera. The first data acquisition device may include: a first laser and a first camera, and the second data acquisition device may include: a second laser and a second camera.

[0076] The first camera and the first laser can be mounted at the same angle. The mounting angle of the first camera can be the angle between the optical axis of the first camera and the vehicle mounting reference surface, and the mounting angle of the first laser can be the angle between the output optical axis of the first laser and the vehicle mounting reference surface. The second camera and the second laser can also be mounted at the same angle. The mounting angle of the second camera can be the angle between the optical axis of the second camera and the vehicle mounting reference surface, and the mounting angle of the second laser can be the angle between the output optical axis of the second laser and the vehicle mounting reference surface. For example, the mounting angle could be 15 degrees.

[0077] The acquisition of the first offset data by operation S201 is achieved through the following operations A1-A2:

[0078] In operation A1: the first image of the contact rail illuminated by the first laser is acquired using the first camera.

[0079] In operation A2: Based on the line laser on the contact rail in the first image, determine the first offset data of the contact rail relative to the running rail in the world coordinate system on the cross section perpendicular to the direction of rail extension.

[0080] In one embodiment of this application, see [link to embodiment]. Figure 3A flowchart of a method for determining a first offset parameter is provided. The above operation A2 can be implemented by the following operations S301-S304.

[0081] In operation S301: the first image is subjected to distortion correction processing to obtain the corrected image.

[0082] In operation S302: using the line laser on the contact rail in the first image, determine the key point pixel coordinates of the key points on the contact rail in the correction image.

[0083] Operation S302 can be achieved through the following operations C1-C3.

[0084] In operation C1: the contact rail region is extracted from the calibration image to obtain the contact rail image.

[0085] In one implementation, the vibration compensation device can use a convolutional neural network to extract the contact rail region from the correction image, thereby obtaining the contact rail region, which may include the side surface region of the contact rail and the bottom surface region of the contact rail region.

[0086] In operation C2: Based on the pixel coordinates of the laser lines in the contact rail image, the three-dimensional model of the contact rail is reconstructed to obtain the three-dimensional model of the contact rail in the world coordinate system.

[0087] In one implementation, the vibration compensation device can determine the three-dimensional point cloud data of the contact rail based on the mapping relationship between the pixel coordinates in the camera coordinate system calibrated by the line laser scanning triangulation method and the three-dimensional spatial coordinates in the world coordinate system. Based on the three-dimensional point cloud data, the three-dimensional model of the contact rail is reconstructed to obtain the three-dimensional model of the contact rail in the world coordinate system.

[0088] In operation C3: Based on the 3D coordinates of the contact rail key points where the calibration area intersects with the laser line in the 3D model of the contact rail, the correspondence between the pre-calibrated 3D coordinates and the sub-pixel coordinates is determined, and the sub-pixel coordinates of the contact rail key points in the calibration image are determined.

[0089] In the 3D model of the contact rail, the calibration area can be the side area and / or the bottom area of ​​the contact rail. The key point of the contact rail where the calibration area intersects with the laser line can be a randomly selected point within the intersection of the calibration area and the laser line.

[0090] In one implementation, the vibration compensation device can convert the three-dimensional coordinates of the contact rail key points into sub-pixel coordinates of the contact rail key points in the correction image based on a pre-calibrated correspondence between three-dimensional coordinates and sub-pixel coordinates.

[0091] The correspondence between pre-calibrated 3D coordinates and sub-pixel coordinates can be calibrated using a dot array.

[0092] It is evident that by pre-calibrating the correspondence between the three-dimensional spatial coordinates of the same position on the contact rail and the pixel coordinates in the captured contact rail image, a sub-pixel level correspondence between pixel coordinates and three-dimensional spatial coordinates can be established. Thus, by using the positions of key points on the model reconstructed from the three-dimensional model and the aforementioned correspondence to calculate the sub-pixel level key point pixel coordinates, the pixel coordinates of key points on the contact rail in the calibration image become more accurate. Furthermore, based on the sub-pixel level key point pixel coordinates, a more accurate three-dimensional position of the contact rail in the world coordinate system can be determined, leading to a more accurate determination of the first offset data and providing reliable data support for subsequent calculations, thereby achieving more accurate vibration compensation.

[0093] In operation S303: Based on the key point pixel coordinates, determine the three-dimensional spatial offset between the standard positions of the contact rails in the camera coordinate system.

[0094] The three-dimensional spatial offset can include: the offset of the key point's three-dimensional spatial position relative to the standard position of the contact rail in the transverse direction of the track, and the offset of the key point's three-dimensional spatial position relative to the standard position of the contact rail in the vertical direction of the track. The vertical direction of the track is the direction of the perpendicular line between the contact rail and the track plane in the cross-section. Alternatively, the offset in the vertical direction of the track can be recorded by recording the standard length of the perpendicular line between the standard position of the contact rail and the track plane in the aforementioned cross-section, and the length of the perpendicular line between the position of the contact rail and the track plane in the aforementioned cross-section.

[0095] In one implementation, the vibration compensation device can perform second-order polynomial fitting to obtain the correspondence between pixel coordinates and three-dimensional spatial offset based on the offset of the three-dimensional spatial offset corresponding to multiple key pixel coordinates on the contact rail in the lateral direction of the rail and the offset of the three-dimensional spatial offset corresponding to the multiple key pixel coordinates in the vertical direction of the rail. Based on the key pixel coordinates and the correspondence between the pixel coordinates and the three-dimensional spatial offset, the three-dimensional spatial offset between the standard positions of the contact rail and the contact rail in the camera coordinate system is determined.

[0096] In another implementation, the vibration compensation device can determine the three-dimensional spatial position of the key point based on the pre-calibrated intrinsic and extrinsic parameters of the first camera and the correspondence between the pre-calibrated ray of the first camera's optical axis and the laser light plane of the first laser in space, and determine the positional offset between the three-dimensional spatial position of the key point and the pre-determined standard position of the contact rail as the three-dimensional spatial offset.

[0097] In operation S304: Based on the three-dimensional spatial offset, the geometric parameters of the contact rail, and the depth information of the corrected image, determine the first offset data of the contact rail relative to the running rail on the cross section perpendicular to the extension direction of the track in the world coordinate system.

[0098] In one implementation, the vibration compensation device can determine the correction amount of the three-dimensional spatial offset based on the geometric parameters of the contact rail and the depth information of the correction image, and use the correction amount to correct the three-dimensional spatial offset to determine the first offset data of the contact rail relative to the running rail on the cross section perpendicular to the extension direction of the track in the world coordinate system.

[0099] For example, the deformation of the contact rail can be determined based on its geometric parameters, the correction amount for the three-dimensional spatial offset can be determined based on the deformation amount, the vertical height of the contact rail relative to the running rail on the cross section perpendicular to the track extension direction can be determined based on the depth information of the corrected image, and the correction amount for the offset of the three-dimensional spatial position of the key point in the three-dimensional spatial offset relative to the standard position of the contact rail in the vertical direction of the track can be determined based on the vertical height.

[0100] As can be seen from the above, in the process of acquiring the first offset data, the first image can be accurately acquired by combining the tilt angle of the laser and the shooting angle of the camera. The pixel coordinates of the laser line can be accurately determined in the three-dimensional space of the key points on the contact rail in the world coordinate system by using perspective projection technology. The pixel coordinates are accurately converted into positions in actual space by mathematical model, which helps to calculate the lateral offset of the contact rail in the actual track system and the vertical height of the contact rail from the track plane of the running rail. The geometric parameters of the contact rail in the actual environment are accurately determined, providing reliable data support for subsequent calculations. It can more accurately evaluate the error caused by the vibration of the vehicle body to the data acquisition equipment, thereby improving the accuracy of the determined force-displacement relationship, more accurately determining the motion displacement of the vehicle body at the time point to be measured, obtaining more accurate vehicle body vibration compensation parameters, and thus more accurately performing vibration compensation on the train at the time point to be measured.

[0101] The following explains how to perform distortion correction using operation S301.

[0102] In one embodiment of this application, the vibration compensation device can perform distortion correction through the following operations D1-D3.

[0103] In operation D1: the first image is preprocessed to obtain a preprocessed image.

[0104] In one implementation, the vibration compensation device downsamples and filters the first image to obtain a lower-layer image. The lower-layer image is then interpolated and dilated to obtain a dilated image. Based on the difference in pixel values ​​between the lower-layer image and the dilated image, a multi-detail layer image is determined. The multi-detail layer images are fused to preserve detail information in each detail layer, resulting in a detail fusion result. Image reconstruction is performed based on the detail fusion result and the multi-detail layer images to obtain a reconstructed image. Adaptive illumination compensation processing is applied to the reconstructed image to obtain an illumination-compensated image. Connected regions in the illumination-compensated image are marked to obtain a pre-processed image.

[0105] The following explains how to fuse images from multiple detail layers.

[0106] In one approach, for each layer of a multi-detail image, the vibration compensation device calculates the local variance and gradient magnitude of the neighborhood of each pixel in the image. Based on the variance weight corresponding to the local variance and the gradient weight corresponding to the gradient magnitude, the fusion weight of the pixel neighborhood to which the local variance and gradient magnitude belong is determined. The multi-detail image is then weighted and fused according to the fusion weight of the pixel neighborhood in each layer of the multi-detail image to obtain the detail fusion result.

[0107] For example, the pixel neighborhood can be a 3×3 pixel area, or a 5×5 pixel area, and so on.

[0108] In this way, by extracting information from different detail levels of the first image and fusing multiple detail level images for image reconstruction, the reconstructed image can retain more detail information of the contact rail, making the contact rail information in the image clearer. This provides more accurate image information support for subsequent edge extraction and distortion correction, making the corrected image more accurate, and thus obtaining more accurate first offset data.

[0109] The following explains the methods for image reconstruction.

[0110] In one approach, for each layer of a multi-detail image, the detail layer image is upsampled, and the upsampled image is added pixel-by-pixel to the detail fusion result of the previous layer, until pixel-by-pixel addition is completed for each layer of the multi-detail image, resulting in a reconstructed image with the same size as the original image. For example, the upsampling factor can be 2x.

[0111] The following section explains how to perform adaptive illumination compensation processing on the reconstructed image to obtain an illumination-compensated image.

[0112] In one approach, the vibration compensation device extracts the background from the reconstructed image to obtain an image of the target object after removing the background from the reconstructed image, and then performs M illumination compensation processes. Here, M is an integer greater than or equal to 1.

[0113] Vibration compensation equipment can extract the background from the reconstructed image to obtain a background estimation image. Subtracting the background estimation image from the reconstructed image yields the target object image after removing the background from the reconstructed image.

[0114] For example, background extraction of a reconstructed image may include the following steps: applying a 7×7 median filter to the distorted fusion image to remove noise and obtain a smooth image; using an erosion-dilation method, selecting a rectangle with a 5×5 structuring element to perform morphological opening on the smooth image to remove bright laser stripes and obtain a background estimation image; and finally, removing the background from the target object image in the reconstructed image.

[0115] The following explains the illumination compensation process.

[0116] The m-th illumination compensation process in the M-th illumination compensation process includes: determining a first non-uniformity coefficient characterizing the degree of grayscale value non-uniformity in the target object image; performing illumination compensation on the target object image based on the m-th illumination compensation coefficient characterizing the illumination supplement intensity to obtain a preliminary illumination-compensated image, where m = 1, 2, ..., M; determining a second non-uniformity coefficient characterizing the degree of grayscale value non-uniformity in the preliminary illumination-compensated image; if the preliminary illumination-compensated image meets the illumination compensation requirements based on the first and second non-uniformity coefficients, then the m-th illumination compensation process is completed, in which case the value of M is equal to the value of m, and the preliminary illumination-compensated image is determined as the illumination-compensated image; if the preliminary illumination-compensated image does not meet the illumination compensation requirements based on the first and second non-uniformity coefficients, then the (m+1)-th illumination compensation coefficient is determined, and the (m+1)-th compensation process is performed.

[0117] See Figure 4 A flowchart of a lighting compensation processing method is provided. The m-th lighting compensation processing includes the following operations S401-S406.

[0118] In operation S401: Determine the first non-uniformity coefficient that characterizes the degree of non-uniformity of gray values ​​in the target object image.

[0119] In one implementation, the vibration compensation device can calculate the standard deviation and mean gray level of each sub-region in the target object image, and calculate the average of the ratios of the standard deviation and mean gray level of each sub-region as a first non-uniformity coefficient. For example, the size of a sub-region can be a 16×16 pixel area.

[0120] In operation S402: Based on the illumination compensation coefficient of the mth iteration representing the illumination supplement intensity, illumination compensation is performed on the target object image to obtain a preliminary illumination-supplemented image.

[0121] In operation S403: Determine the second non-uniformity coefficient, which characterizes the degree of non-uniformity of gray values ​​in the preliminary optically compensated image.

[0122] The method of determining the second non-uniformity coefficient in operation S403 is similar to that of determining the first non-uniformity coefficient. The difference lies in the different names of the second and first non-uniformity coefficients, and the different names of the preliminary light-compensated image and the target object image, which will not be described in detail here.

[0123] In operation S404: Based on the first non-uniformity coefficient and the second non-uniformity coefficient, determine whether the preliminary light-compensated image meets the illumination compensation requirements.

[0124] If the result indicates that the illumination compensation requirement is met, then operation S405 is executed; if the result indicates that the illumination compensation requirement is not met, then operation S406 is executed.

[0125] In one implementation, if the ratio of the first non-uniformity coefficient to the second non-uniformity coefficient is less than the first non-uniformity threshold, then it is determined whether the preliminary light-compensated image meets the illumination compensation requirements.

[0126] If the ratio of the first non-uniformity coefficient to the second non-uniformity coefficient is greater than or equal to the first non-uniformity threshold and less than or equal to the second non-uniformity threshold, then it is determined that the preliminary light compensation image does not meet the illumination compensation requirements and the compensation is insufficient. In operation S406, the illumination compensation coefficient at the m-th time can be increased by a preset adjustment step size to obtain the illumination compensation coefficient at the (m+1)-th time.

[0127] If the ratio of the first non-uniformity coefficient to the second non-uniformity coefficient is greater than or equal to the second non-uniformity threshold, it is determined that the preliminary light compensation image does not meet the illumination compensation requirements and is over-compensated. In operation S406, the illumination compensation coefficient at the m-th time can be reduced by a preset adjustment step size to obtain the illumination compensation coefficient at the (m+1)-th time.

[0128] In operation S405: complete the m-th illumination compensation process and determine the preliminary illumination compensation image as the illumination compensation image.

[0129] In this case, the value of M is equal to the value of m. The value of M can also be set to a maximum value. If m=M when the value of M is set to a maximum value, then the preliminary light-compensated image can be determined as the illumination-compensated image after operation S402.

[0130] In operation S406: Determine the illumination compensation coefficient for the (m+1)th time and perform the compensation process for the (m+1)th time.

[0131] After operating S406, you can return to operating S402 to perform the (m+1)th compensation process.

[0132] By iteratively compensating for illumination until the image meets the illumination compensation requirements through the non-uniformity coefficient, the clarity of darker areas in the image is improved. Combined with background extraction, image noise can be reduced, making the contact rail information in the image clearer. Especially in complex real-world application environments, it can effectively remove interference caused by illumination. Using clearer and more accurate images for processing improves the accuracy of the first offset data acquired.

[0133] Furthermore, after illumination compensation, the grayscale value of the illumination-compensated image can be calculated. Based on the calculated grayscale value, the illumination-compensated image is dilated. The image after dilation is segmented to identify defective pixels, and the remaining pixels in the illumination-compensated image are marked to generate connected regions, thus obtaining a preliminary processed image.

[0134] After operation D1 preprocesses the first image to obtain the preprocessed image.

[0135] In operation D2: edge extraction is performed on the preprocessed image to obtain the edge bounding box of the laser on the contact rail in the preprocessed image.

[0136] For example, vibration compensation equipment can use the Canny algorithm (an edge detection algorithm) to extract edges from the preprocessed image, obtaining the edge bounding box of the laser on the contact rail in the preprocessed image.

[0137] In operation D3: the preprocessed image is subjected to N distortion correction processes to obtain the corrected image.

[0138] Where N is an integer greater than or equal to 1.

[0139] The distortion correction process is explained below.

[0140] In one embodiment of this application, the nth distortion correction process includes: deforming the preprocessed image based on the pixel coordinates of the vertices of the nth bounding box to obtain a deformed image, where n = 1, 2, ..., N. The rectangularity of the bounding boxes in the deformed image is determined. The rectangularity characterizes how closely the shape of the bounding box approximates a rectangle. If the rectangularity characterizes the bounding boxes in the deformed image to meet the rectangular requirement, the nth distortion correction process is completed, and the deformed image is determined as the corrected image; if the rectangularity characterizes the bounding boxes in the deformed image to not meet the rectangular requirement, the nth bounding box in the deformed image is refitted to obtain the (n+1)th bounding box, and the (n+1)th distortion correction process is performed.

[0141] In one embodiment of this application, see Figure 5 A flowchart of a distortion correction method is provided, wherein the nth distortion correction process includes the following operations S501-S506.

[0142] In operation S501: Based on the pixel coordinates of the vertices of the nth bounding box, the preprocessed image is deformed to obtain a deformed image.

[0143] In one implementation, the vibration compensation device determines the camera's tilt feature when capturing the first image based on the pixel coordinates of the vertices of the nth bounding box. Based on the tilt feature, a first deformation correction is performed on the preprocessed image to obtain a preliminary corrected image. Based on the distortion feature of the camera that captured the first image and the camera's intrinsic parameters, a second deformation correction is performed on the preliminary corrected image to achieve deformation removal, resulting in a deformed image.

[0144] In one approach, the vibration compensation device can perform a first deformation correction on the preprocessed image using perspective transformation to obtain a preliminary corrected image, and then perform a second deformation correction on the preliminary corrected image using the camera intrinsic parameter matrix and distortion coefficients to obtain a deformed image.

[0145] The distortion features and intrinsic parameters of the camera acquiring the first image can be pre-calibrated using checkerboard corner point technology. Distortion features are parameters that describe the degree of optical distortion of the camera lens. Distortion features can include radial distortion coefficients and tangential distortion coefficients. Distortion features are affected by the optical structure and focal length of the lens, and are determined by the camera's inherent parameters and its own characteristics.

[0146] The tilt feature can describe the rotation angle and translation of the entire image relative to the track plane coordinate system. The tilt feature is affected by the camera mounting tilt angle and the real-time vibration attitude of the vehicle body.

[0147] As can be seen from the above, the tilt feature quantity is calculated by using the coordinates of the vertices of the laser's outer frame to determine the trapezoidal deformation correction. By correcting the outer frame of the laser, the tilt deformation of the entire preprocessed image is reduced. Furthermore, the distortion feature quantity and camera intrinsic parameters are used to perform distortion removal processing to reduce the distortion caused by bending and rotation, making the outer frame closer to a rectangle. This makes the entire image closer to the original geometry of the contact rail, making the corrected image more accurate, and thus obtaining more accurate first offset data.

[0148] In operation S502: Determine the rectangularity of the bounding box in the deformed image.

[0149] Among them, rectangularity characterizes how closely the shape of the bounding box resembles a rectangle.

[0150] In one implementation, the vibration compensation device can determine the difference between the angles of the four interior corners of the edge box in the deformed image and 90 degrees, and use this difference as the rectangularity of the edge box in the deformed image.

[0151] In operation S503: Determine whether the bounding box in the deformed image meets the rectangularity requirement.

[0152] In one implementation, if the difference between the angles of the four interior corners of the bounding box in the deformed image and 90 degrees is less than or equal to 1 degree, then the rectangularity is determined to indicate that the bounding box in the deformed image meets the rectangular requirement. If the difference between the angles of the four interior corners of the bounding box in the deformed image and 90 degrees is greater than 1 degree, then the rectangularity is determined to indicate that the bounding box in the deformed image does not meet the rectangular requirement.

[0153] If the bounding box in the rectangularity-represented deformed image meets the rectangularity requirement, then operation S504 is executed; if the bounding box in the rectangularity-represented deformed image meets the rectangularity requirement, then operation S505 is executed.

[0154] In operation S504: complete the nth distortion correction process and determine the deformed image as the corrected image.

[0155] In this case, the value of N is equal to the value of n. The value of N can also be set to a maximum value. If the value of N is set to a maximum value and n=N, then the deformed image can be determined as a lighting-compensated image after operation S501.

[0156] In operation S505: the edge box of the nth time in the deformed image is refitted to obtain the edge box of the (n+1)th time, and the distortion correction process is performed for the (n+1)th time.

[0157] In one implementation, the vibration compensation device can use RANSAC (Random Sample Consensus) to refit the edge box of the nth time in the deformed image to obtain the edge box of the (n+1)th time, and perform the distortion correction process of the (n+1)th time.

[0158] Furthermore, the vibration compensation device can detect residual radial and tangential distortion in the deformed image. If residual radial distortion exists, the radial distortion coefficient is adjusted; if residual tangential distortion exists, the tangential distortion coefficient is adjusted.

[0159] In this way, by judging the rectangularity of the bounding box in the deformed image, the distortion correction effect of the entire deformed image can accurately reflect whether the image distortion in the preprocessed image has been effectively corrected, improve the accuracy of the contact rail shape in the corrected image, and thus improve the accuracy of the contact rail geometric information in the image, making the subsequent data processing more reliable. Especially in complex practical application environments, it can effectively remove the interference caused by distortion, and thus obtain more accurate first offset data.

[0160] The acquisition of the second offset data by operation S201 is achieved through the following operations B1-B3:

[0161] In operation B1: Multiple second images of the travel track illuminated by the second laser are acquired using the second camera.

[0162] In operation B2: Based on the pixel coordinates of the line lasers on the travel track in multiple second images, determine the laser positions of the corresponding line lasers in the world coordinate system in multiple second images.

[0163] In one implementation, the pixel coordinates of the line lasers on the travel path in multiple second images can be transformed into the laser positions of the corresponding line lasers in the world coordinate system by using the pre-calibrated intrinsic and extrinsic parameters of the second camera that acquires the second images. For example, the second camera can be calibrated using the Zhang Zhengyou calibration method to obtain its intrinsic and extrinsic parameters.

[0164] In operation B3: Based on the laser positions corresponding to multiple second images, determine the second offset data of the relative position between the vehicle body and the running track.

[0165] In one implementation, the vibration compensation device can determine the second offset data of the relative position between the vehicle body and the running track by collecting the difference between the laser positions of two consecutive second images corresponding to the line laser in the world coordinate system.

[0166] When the vehicle body vibrates, the position where the line laser intersects with the surface of the running track will change, and the position of the laser stripe will also shift accordingly. By observing the shift of the laser position of the line laser corresponding to multiple second images in the world coordinate system, the dynamic position change of the vehicle body on the track plane can be reflected. In this way, the second offset data of the relative position between the vehicle body and the running track can be determined based on the laser position corresponding to multiple second images.

[0167] By acquiring a first image of the laser line on the contact rail, performing image processing on the first image to determine the first offset data, and acquiring a second image of the laser line on the running rail, the second offset data is obtained from the deviation between the laser positions corresponding to multiple second images through image processing. This data acquisition process can extract high-precision first and second offset data, minimize the impact of vibration on the data acquisition equipment during data acquisition, provide reliable data support for subsequent calculations, more accurately assess the error caused by vehicle vibration on the data acquisition equipment, thereby improving the accuracy of the determined force-displacement relationship, more accurately determining the vehicle's motion displacement at the measured time point, obtaining more accurate vehicle vibration compensation parameters, and thus more accurately performing vibration compensation on the train at the measured time point.

[0168] The following explains how to construct the pose change relationship.

[0169] The process of obtaining a new pose change relationship can be executed periodically. After obtaining the new pose change relationship, operation S201 can determine the vehicle vibration pose at the time point to be measured based on the new pose change relationship.

[0170] According to an embodiment of this application, see Figure 6 A flowchart of a method for constructing pose change relationships is provided. The method includes the following operations S601-S603.

[0171] In operation S601: Based on the first offset number, the second offset data, and the error law characterizing the actual position of the contact rail and the observed position of the contact rail, determine the measurement error caused by the vehicle vibration to the position parameter detection of the contact rail.

[0172] The actual position of the contact rail is the position of the contact rail when only the vehicle body is affected by vibration, while the observed position of the contact rail is the position of the contact rail acquired by the data acquisition equipment when both the vehicle body and the data acquisition equipment are affected by vehicle body vibration.

[0173] In one implementation, the lateral measurement error in the transverse direction and the vertical measurement error in the vertical direction caused by the vehicle vibration on the position parameter detection of the contact rail can be determined based on a preset error law representing the error between the actual position and the observed position of the contact rail, using the first offset number and the second offset data.

[0174] For example, the lateral measurement error can be determined based on the following expression:

[0175] ;

[0176] in, Indicates lateral measurement error. This indicates the offset of the second offset data in the lateral direction of the track. This indicates the offset of the first offset data in the vertical direction of the track. This indicates the camera's position in the vertical direction of the track.

[0177] The vertical measurement error can be determined based on the following expression:

[0178] ;

[0179] in, Indicates vertical measurement error. This indicates the offset of the second offset data in the vertical direction of the track. This indicates the offset of the first offset data in the lateral direction of the track. This indicates the position of the camera in the horizontal direction of the track.

[0180] In operation S602: Based on multiple sets of first offset data and the measurement errors corresponding to multiple sets of first offset data, determine multiple sets of vehicle body vibration poses caused by vibration when the data acquisition device generates measurement errors due to vibration.

[0181] In one implementation, operation S602 can be implemented through the following operations E1-E2.

[0182] In operation E1: multiple sets of first offset data and the corresponding measurement errors are modeled as random variables.

[0183] The random variable represents the uncertainty of the first offset data collected under the conditions of vehicle body vibration and data acquisition equipment vibration.

[0184] In one approach, for each set of first offset data, the time point of the first offset data can be recorded. The time point of the measurement error is the same as the time point of the first offset data corresponding to the measurement error. The vibration compensation device models the first offset data and the measurement error corresponding to the first offset data at each time point as random variables in the form of a six-dimensional vector.

[0185] In operation E2: obtain the pose representation of the random variables corresponding to multiple sets of first offset data, as multiple sets of vehicle body vibration poses generated by the vehicle body under the condition of measurement error.

[0186] In one approach, for each set of first offset data corresponding to a random variable, the random variable can be parameterized as a six-dimensional vector of pose increment. This six-dimensional vector is then mapped to a Lie algebra linear space via an exponential mapping to obtain the pose transformation matrix corresponding to that set of first offset data, which serves as the vehicle vibration pose. The time point of the pose transformation matrix corresponding to each set of first offset data is the same as the time point of the first offset data itself, thus obtaining multiple sets of vehicle vibration poses at different time points.

[0187] In this way, by modeling the measurement error as a random variable, the ability of the constructed pose change relationship to quantify vibration data with uncertainty can be improved. Furthermore, by using such a pose change relationship, the vibration pose of the vehicle body caused by vibration at the time point under test can be determined more accurately, thereby improving the accuracy of vibration compensation for the train.

[0188] In operating S603: Construct pose change relationships based on multiple sets of vehicle body vibration poses.

[0189] In one implementation, the pose change relationship is obtained by constructing multiple sets of vehicle body vibration poses into a time-conditional Gaussian mixture model.

[0190] As can be seen from the above, by combining multiple sets of first offset data and multiple sets of second offset data, the measurement error of the contact rail parameters collected by the acquisition equipment due to vehicle vibration can be evaluated and calculated in real time. Based on the measurement error caused by vibration, the vehicle vibration pose caused by vibration at the same time point can be evaluated, which can more accurately determine the vehicle pose caused by vibration. In this way, a more comprehensive dynamic pose evaluation framework can be constructed, and the accuracy of the constructed pose change relationship can be improved.

[0191] According to an embodiment of this application, the vibration compensation device can also collect the vehicle body vibration pose at multiple time points under the influence of vibration based on sensors installed on the train, and construct a pose change relationship based on multiple sets of vehicle body vibration poses.

[0192] The method for determining motion displacement is explained below.

[0193] According to an embodiment of this application, the above operation S202 can be implemented by the following operations E1-E3:

[0194] In operation E1: bandpass filtering is performed on the vehicle body vibration pose to obtain the low-frequency and high-frequency components of the vibration.

[0195] In one implementation, a time window of a preset time length is determined with the time point to be measured corresponding to the vehicle body vibration pose as the center. Based on the pose change relationship, the poses corresponding to multiple time points in the time window are determined. The vehicle body vibration poses at multiple time points are bandpass filtered to represent the low-frequency and high-frequency components of the vibration at the time point to be measured.

[0196] The low-frequency components of vibration include vibrations caused by track geometric deviations due to installation errors and wear, primarily affecting the steady-state accuracy of vibration compensation. For example, the frequency of the low-frequency components can be less than 1 Hz. The high-frequency components of vibration include vehicle body vibrations caused by wheel-rail impacts and suspension system vibrations, primarily affecting the dynamic response speed of compensation. For example, the frequency of the high-frequency components can be greater than or equal to 1 Hz and less than 20 Hz.

[0197] In operation E2: based on the low-frequency and high-frequency components of vibration, the desired pose is converted into an equivalent excitation quantity characterizing the vibration intensity experienced by the vehicle body.

[0198] Specifically, the equivalent excitation quantity corresponding to multiple sets of first offset data collected at multiple acquisition times can characterize the effect of the equivalent excitation quantity of the acquisition time to which the first offset data belongs on the vehicle displacement at the next acquisition time. In other words, the equivalent excitation quantity can characterize the effect of the energy carried by the vibration on the vehicle displacement at the next time point to which it belongs.

[0199] The low-frequency components of vibration have a relatively small impact on the equivalent excitation value, and this impact changes slowly over time. Conversely, the high-frequency components of vibration have a significant impact on the equivalent excitation value, and the low-frequency components' impact changes rapidly over time.

[0200] In one implementation, the vibration compensation device can convert the desired pose into an equivalent excitation quantity characterizing the vibration intensity experienced by the vehicle body based on the low-frequency components and high-frequency components of the vibration, the dynamic transfer function matrix corresponding to the low-frequency components of the pre-calibrated vehicle suspension system, and the dynamic transfer function matrix corresponding to the high-frequency components.

[0201] For example, the equivalent excitation can be determined using the following expression:

[0202] ;

[0203] in, This represents the equivalent stimulus at the time point to be measured. This represents the dynamic transfer function matrix corresponding to the low-frequency components. This represents the low-frequency components at the time point to be measured. This represents the dynamic transfer function matrix corresponding to the high-frequency components. This represents the high-frequency components at the time point to be measured.

[0204] In operation E3: Based on the pre-constructed force-displacement relationship between the equivalent excitation and the motion displacement, determine the motion displacement of the vehicle body relative to the running track at the time point to be measured corresponding to the equivalent excitation.

[0205] The force-displacement relationship is constructed based on multiple motion displacements obtained historically and multiple equivalent excitation quantities corresponding to these motion displacements.

[0206] In one implementation, the equivalent excitation at the time point to be measured is input into a pre-constructed force-displacement relationship to obtain the motion displacement of the vehicle body relative to the running track at the time point to be measured.

[0207] In another implementation, after inputting the equivalent excitation amount of the time point to be measured into the pre-constructed force-displacement relationship to obtain the first motion displacement of the vehicle body relative to the running rail at the time point to be measured, the motion displacement residual value of the time point to be measured can also be obtained. The motion displacement residual value of the time point to be measured is used to correct the first motion displacement of the time point to be measured to obtain the second motion displacement, and the second motion displacement is determined as the motion displacement of the time point to be measured.

[0208] The residual values ​​of motion displacement at the time point to be measured can be obtained from a pre-constructed residual prediction model. The time point to be measured is input into the residual prediction model to obtain the residual values ​​of motion displacement. For example, the residual prediction model can be constructed based on historical residual values ​​using a gray prediction model. For instance, historical residual values ​​can be determined by collecting the actual values ​​of motion displacement at historical time points and the predicted values ​​of motion displacement at historical time points determined by E1-E3 as described above.

[0209] By combining multiple sets of first offset data, the measurement error of the contact rail parameters acquired by the acquisition equipment due to vehicle vibration can be evaluated and calculated in real time, providing a more comprehensive dynamic evaluation framework. At the same time, by using the generation of pose change matrix and bandpass filtering technology, the low-frequency and high-frequency components of vibration are analyzed, so that the calculated force-displacement relationship can effectively determine the motion displacement in various vibration modes, and accurately predict the motion displacement in dynamic environments, providing high-precision motion displacement for the vibration compensation control process.

[0210] According to an embodiment of this application, the force-displacement relationship can be achieved by following operations F1-F3.

[0211] In operation F1: Based on multiple motion displacements and the multiple equivalent excitation quantities corresponding to the multiple motion displacements, extract the correlation between the equivalent excitation quantities and the motion displacements;

[0212] In one implementation, the vibration compensation device can extract the correlation between the equivalent excitation and the motion displacement based on a preset mathematical model.

[0213] For example, linear or exponential mathematical models can be used to extract the correlation between equivalent excitation and motion displacement.

[0214] In operation F2: Based on the correlation law, determine the influence weight of the equivalent excitation on the motion displacement.

[0215] In one implementation, the vibration compensation device can determine the contribution coefficient of each force dimension component in the correlation law to the displacement change as the influence weight of the equivalent excitation on the motion displacement.

[0216] In operation F3: perform Z relation construction processes.

[0217] Where Z is an integer greater than or equal to 1.

[0218] In one implementation, see [link to implementation details]. Figure 7 A flowchart illustrating a method for constructing force-displacement relationships is provided. The z-th relationship construction process includes the following operations S701-S705, where z = 1, 2, ..., Z.

[0219] In operation S701: Based on the equivalent incentive amount, correlation law and the influence weight of the z-th time, candidate correspondence is constructed with the equivalent incentive amount as the independent variable and the motion displacement as the dependent variable.

[0220] When z=1, the influence weight of the z-th operation can be the influence weight determined by operation F2.

[0221] In one implementation, the vibration compensation device can adopt a second-order exponential decay fitting model. The least squares fitting method is used to construct candidate correspondences based on the equivalent excitation, correlation law, and the influence weight of the z-th time, with the equivalent excitation as the independent variable and the motion displacement as the dependent variable.

[0222] For example, the candidate correspondence can be represented by the following expression:

[0223] ;

[0224] in, The displacement represents the motion at the time point to be measured, and A represents the amplitude vector. Indicates the attenuation coefficient. Indicates the natural angular frequency of vibration. The initial phase is represented by B, and the DC component is represented by B. The amplitude vector is determined based on the high-frequency components of the equivalent excitation and the dynamic gain in the influence weights. The natural angular frequency of the vibration is determined based on the dominant frequency of the equivalent excitation, and the initial phase is determined based on the phase of the equivalent excitation at its initial time point.

[0225] In addition, the equivalent stimulus used in operating S701 may include a known equivalent stimulus for a continuous historical time window.

[0226] In operation S702: Determine the candidate motion displacement at the test time point based on the candidate correspondence.

[0227] The test time point can be any time point after the time point to which the equivalent stimulus used in operation S701 belongs, except for the time point to which the equivalent stimulus used in operation S701 belongs.

[0228] In one implementation, the vibration compensation device can input the known equivalent excitation at the test time point into the above candidate correspondence to obtain the candidate motion displacement at the test time point.

[0229] In operation S703: Based on the pre-built error prediction model, determine whether the prediction error rate of the candidate motion displacement is higher than the preset error threshold.

[0230] In one implementation, the error prediction model can be constructed using a gray prediction model. For example, a gray prediction model for predicting the error rate can be constructed using the known error rate of a preset number of time points. The test time points are then input into the constructed gray prediction model for predicting the error rate to obtain the prediction error rate of the candidate motion displacement.

[0231] If the prediction error rate of the candidate motion displacement determined based on the pre-built error prediction model is higher than the preset error threshold, operation S704 is executed. If the prediction error rate is less than or equal to the preset error threshold, the z-th relation construction process is completed and operation S705 is executed.

[0232] In operation S704: Based on the correlation rules, determine the adjustment influence weight of the (z+1)th relation and perform the (z+1)th relation construction process.

[0233] In one implementation, the vibration compensation device can update the influence weights based on the prediction error using the recursive least squares method or the gradient descent method to obtain the adjusted influence weights for the (z+1)th time.

[0234] In operation S705: complete the z-th relation construction process and determine the candidate correspondence as a force-displacement relation.

[0235] In this case, the value of Z is equal to the value of z. The value of Z can also be set to a maximum value. If z = Z when the value of Z is set to a maximum value, then after operating S401, the candidate correspondence can be determined as a force-displacement relationship.

[0236] In this way, by evaluating the prediction error rate of the force-displacement relationship through the error prediction model and dynamically adjusting the influence weights, the determined candidate correspondence can be made more suitable for the vibration process of the train within a certain period of time. Furthermore, by dynamically adjusting the displacement change relationship in this way, it can be applied to the vibration process of rail vehicles that has both periodicity and randomness, thereby improving the accuracy of the determined displacement change relationship.

[0237] By determining the equivalent excitation required for vibration to affect the car body under the condition of measurement error by using multiple sets of first offset data and the corresponding measurement errors, the multidimensional characteristics of contact rail geometric errors can be effectively captured. Furthermore, by combining the relative position of the contact rail with the running rail, the obtained force-displacement relationship can reflect the specific impact of car body vibration on the track geometry, including both the contact rail and the running rail. It can take into account track geometric changes and comprehensively reflect various errors caused by vibration, thus providing error data support for car body vibration compensation and improving the accuracy of error modeling. The car body vibration compensation parameters obtained by converting motion displacement using this force-displacement relationship can effectively optimize the vibration compensation effect, reduce the impact of car body vibration on the track, and improve track stability and car body comfort.

[0238] In the embodiments of this application, the above step S203 of converting vehicle vibration compensation parameters can be implemented in the following way: establishing a mapping relationship between the force and displacement transmission characteristics of the vehicle active suspension system to construct a parameter conversion relationship between motion displacement and vehicle vibration compensation parameters; based on the parameter conversion relationship, converting motion displacement into vehicle vibration compensation parameters.

[0239] For example, the expression for establishing the mapping relationship between the force and displacement transmission characteristics of a vehicle's active suspension system can be:

[0240] ;

[0241] in, This indicates the vehicle body vibration compensation parameters. Indicates displacement. Represents the proportional gain matrix. Let represent the differential gain matrix, where the proportional gain matrix and the differential gain matrix are pre-constructed based on the force and displacement transmission characteristics of the vehicle's active suspension system.

[0242] In this way, by dynamically adjusting the parameter conversion relationship between motion displacement and vehicle vibration compensation parameters, the vehicle vibration compensation parameters can be made more suitable for the vibration state of the dynamically operating train, providing more accurate vehicle vibration compensation parameters for the vibration compensation control process, and thus enabling more accurate vibration compensation for the train.

[0243] Corresponding to the vibration compensation method described above, this application also provides a vibration compensation device.

[0244] In one embodiment of this application, see Figure 8A schematic diagram of a vibration compensation device is provided. The vibration compensation device 800 includes: a vehicle body vibration posture determination module 801, a motion displacement determination module 802, a compensation parameter conversion module 803, and a vibration compensation module 804.

[0245] The vehicle body vibration pose determination module 801 is used to determine the vehicle body vibration pose at the time point to be measured by vibration based on the pre-constructed pose change relationship of the vehicle body under vibration influence over time. The pose change relationship is constructed based on multiple sets of first offset data and multiple sets of second offset data collected by the data acquisition equipment installed on the vehicle body during the train's operation. The multiple sets of first offset data and multiple sets of second offset data are respectively: multiple sets of offset data on the relative position between the train's contact rail and running rail changing over time under the influence of vehicle body vibration and data acquisition equipment vibration, and multiple sets of offset data on the relative position between the vehicle body and running rail changing over time.

[0246] The motion displacement determination module 802 is used to determine the motion displacement of the vehicle body relative to the running track at the time point to be measured based on the vehicle body vibration posture.

[0247] The compensation parameter conversion module 803 is used to convert motion displacement into vehicle vibration compensation parameters.

[0248] The vibration compensation module 804 is used to perform vibration compensation on the train based on the vehicle body vibration compensation parameters.

[0249] According to an embodiment of this application, the data acquisition device includes: a first laser, a first camera, a second laser, and a second camera; wherein, the offset data acquisition module acquires first offset data in the following manner: using the first camera to acquire a first image of the contact rail irradiated by the first laser; based on the line laser on the contact rail in the first image, determining the first offset data of the contact rail relative to the running rail in a cross section perpendicular to the track extension direction in the world coordinate system; wherein, the offset data acquisition module acquires second offset data in the following manner: using the second camera to acquire multiple second images of the running rail irradiated by the second laser; based on the pixel coordinates of the line laser on the running rail in the multiple second images, determining the laser position of the line laser corresponding to the multiple second images in the world coordinate system; and determining the second offset data of the relative position between the vehicle body and the running rail based on the laser positions corresponding to the multiple second images.

[0250] According to an embodiment of this application, the offset data acquisition module is used to perform distortion correction processing on the first image to obtain a corrected image; using the line laser on the contact rail in the first image, the key point pixel coordinates of the key points on the contact rail in the corrected image are determined; based on the key point pixel coordinates, the three-dimensional spatial offset between the standard positions of the contact rail and the contact rail in the camera coordinate system is determined; based on the three-dimensional spatial offset, the geometric parameters of the contact rail, and the depth information of the corrected image, the first offset data of the contact rail relative to the running rail on the cross section perpendicular to the track extension direction in the world coordinate system is determined.

[0251] According to an embodiment of this application, the offset data acquisition module is used to preprocess the first image to obtain a preprocessed image; to extract edges from the preprocessed image to obtain the edge box of the laser on the contact rail in the preprocessed image; to perform N distortion correction processes on the preprocessed image to obtain a corrected image, where N is an integer greater than or equal to 1, and the nth distortion correction process includes: based on the pixel coordinates of the vertices of the nth edge box, to perform deformation processing on the preprocessed image to obtain a deformed image, where n = 1, 2...N; to determine the rectangularity of the edge box in the deformed image, where the rectangularity characterizes the degree to which the shape of the edge box is close to a rectangle; if the rectangularity characterizes the edge box in the deformed image to meet the rectangle requirement, the nth distortion correction process is completed, and the deformed image is determined as the corrected image; if the rectangularity characterizes the edge box in the deformed image to not meet the rectangle requirement, the nth edge box in the deformed image is refitted to obtain the (n+1)th edge box, and the (n+1)th distortion correction process is performed.

[0252] According to an embodiment of this application, the offset data acquisition module is used to determine the tilt feature of the camera when capturing the first image based on the pixel coordinates of the vertices of the edge box in the nth iteration; based on the tilt feature, the preprocessed image is subjected to a first deformation correction to obtain a preliminary corrected image; based on the distortion feature of the camera that acquired the first image and the camera intrinsic parameters, the preliminary corrected image is subjected to a second deformation correction to achieve deformation removal processing and obtain a deformed image.

[0253] According to an embodiment of this application, the offset data acquisition module is used to downsample and filter the first image to obtain a lower layer image; to interpolate and dilate the lower layer image to obtain a dilated image; to determine a multi-detail layer image based on the difference in pixel values ​​between the lower layer image and the dilated image; to fuse the multi-detail layer images to retain detail information in each detail layer to obtain a detail fusion result; to reconstruct the image based on the detail fusion result and the multi-detail layer images to obtain a reconstructed image; to perform adaptive illumination compensation processing on the reconstructed image to obtain an illumination-compensated image; and to mark connected regions in the illumination-compensated image to obtain a preliminary processed image.

[0254] According to an embodiment of this application, the offset data acquisition module is used to extract the background from the reconstructed image to obtain a target object image after removing the background from the reconstructed image; perform M illumination compensation processes, where M is an integer greater than or equal to 1, and the m-th illumination compensation process includes: determining a first non-uniformity coefficient characterizing the degree of grayscale value non-uniformity in the target object image; performing illumination compensation on the target object image based on the m-th illumination compensation coefficient characterizing the illumination supplement intensity to obtain a preliminary illumination supplement image, where m = 1, 2, ..., M; determining a second non-uniformity coefficient characterizing the degree of grayscale value non-uniformity in the preliminary illumination supplement image; if the preliminary illumination supplement image meets the illumination compensation requirements based on the first and second non-uniformity coefficients, then the m-th illumination compensation process is completed, and the preliminary illumination supplement image is determined as the illumination compensation image; if the preliminary illumination supplement image does not meet the illumination compensation requirements based on the first and second non-uniformity coefficients, then the (m+1)-th illumination compensation coefficient is determined, and the (m+1)-th compensation process is performed.

[0255] According to an embodiment of this application, the offset data acquisition module is used to extract the contact rail region in the calibration image to obtain a contact rail image; based on the pixel coordinates of the laser line in the contact rail image, the three-dimensional model of the contact rail is reconstructed to obtain a three-dimensional model of the contact rail in the world coordinate system; based on the three-dimensional coordinates of the contact rail key points where the calibration area intersects with the laser line in the three-dimensional model of the contact rail, the correspondence between the pre-calibrated three-dimensional coordinates and the sub-pixel coordinates, the sub-pixel level contact rail pixel coordinates of the contact rail key points in the calibration image are determined.

[0256] According to an embodiment of this application, the above-mentioned device further includes a change relationship construction module for determining the measurement error caused by vehicle vibration to the detection of the position parameters of the contact rail based on a first offset number, second offset data, and an error law characterizing the actual position of the contact rail and the observed position of the contact rail. The actual position of the contact rail is the position of the contact rail under conditions where only vehicle vibration affects it, and the observed position of the contact rail is the position of the contact rail acquired by the data acquisition device when both the vehicle body and the data acquisition device are affected by vehicle vibration. Based on multiple sets of first offset data and the corresponding measurement errors, multiple sets of vehicle body vibration poses caused by vibration are determined when the vibration causes measurement errors in the data acquisition device. A pose change relationship is constructed based on these multiple sets of vehicle body vibration poses.

[0257] According to an embodiment of this application, the change relationship construction module is used to model multiple sets of first offset data and the measurement error corresponding to multiple sets of first offset data as random variables, wherein the random variables represent the uncertainty of the first offset data collected under the influence of vehicle body vibration and the influence of vehicle body vibration on the data acquisition device; and to obtain the pose representation of the random variables corresponding to multiple sets of first offset data as multiple sets of vehicle body vibration poses generated by the vehicle body under the condition of measurement error.

[0258] According to an embodiment of this application, the motion displacement determination module 802 is used to perform bandpass filtering on the vibration pose of the vehicle body to obtain the low-frequency and high-frequency components of the vibration; based on the low-frequency and high-frequency components of the vibration, the desired pose is converted into an equivalent excitation quantity characterizing the vibration intensity experienced by the vehicle body; according to the pre-constructed force-displacement relationship between the equivalent excitation quantity and the motion displacement, the motion displacement of the vehicle body relative to the running track at the time point to be measured corresponding to the equivalent excitation quantity is determined, wherein the force-displacement relationship is constructed based on multiple historically acquired motion displacements and multiple equivalent excitation quantities corresponding to the multiple motion displacements.

[0259] The motion displacement determination module 802 is used to extract the correlation between equivalent excitation quantities and motion displacements based on multiple motion displacements and their corresponding equivalent excitation quantities; determine the influence weight of equivalent excitation quantities on motion displacements based on the correlation patterns; perform Z-fold relationship construction processing, where Z is an integer greater than or equal to 1, and the z-th relationship construction processing includes: constructing candidate correspondences based on equivalent excitation quantities, correlation patterns, and the z-th influence weights, with equivalent excitation quantities as independent variables and motion displacements as dependent variables, where z = 1, 2, ..., Z; determining candidate motion displacements at test time points based on the candidate correspondences; if the prediction error rate of the candidate motion displacements determined based on the pre-built error prediction model is higher than a preset error threshold, then the z+1-th adjustment influence weight is determined based on the correlation patterns, and the z+1-th relationship construction processing is performed; if the prediction error rate is less than or equal to the preset error threshold, then the z-th relationship construction processing is completed, and the candidate correspondences are determined as force-displacement relationships.

[0260] According to an embodiment of this application, the compensation parameter conversion module 803 is used to establish a mapping relationship between the force and displacement transmission characteristics of the vehicle active suspension system, so as to construct a parameter conversion relationship between motion displacement and vehicle vibration compensation parameters; based on the parameter conversion relationship, the motion displacement is converted into vehicle vibration compensation parameters.

[0261] Any one or more of the modules, submodules, units, and subunits according to the embodiments of this application, or at least part of the functions of any one or more of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to the embodiments of this application can be implemented by dividing them into multiple modules. Any one or more of the modules, submodules, units, and subunits according to the embodiments of this application can be at least partially implemented as hardware circuits, such as field-programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems-on-a-chip, systems-on-a-substrate, systems-on-package, application-specific integrated circuits (ASICs), or implemented by hardware or firmware in any other reasonable manner by integrating or packaging circuits, or implemented in any one of software, hardware, and firmware, or in a suitable combination of any of these. Alternatively, one or more of the modules, submodules, units, and subunits according to the embodiments of this application can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0262] For example, any multiple of the vehicle body vibration pose determination module 801, motion displacement determination module 802, compensation parameter conversion module 803, and vibration compensation module 804 can be combined into one module / unit / subunit, or any one of these modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least part of the functionality of one or more of these modules / units / subunits can be combined with at least part of the functionality of other modules / units / subunits and implemented in one module / unit / subunit. According to embodiments of this application, at least one of the vehicle body vibration pose determination module 801, motion displacement determination module 802, compensation parameter conversion module 803, and vibration compensation module 804 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any appropriate combination of any of these three implementation methods. Alternatively, at least one of the vehicle body vibration pose determination module 801, motion displacement determination module 802, compensation parameter conversion module 803, and vibration compensation module 804 can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0263] Figure 9 A block diagram of an electronic device suitable for implementing the methods described above, according to an embodiment of this application, is illustrated schematically. Figure 9The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0264] like Figure 9 As shown, an electronic device 900 according to an embodiment of this application includes a processor 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage portion 908 into a random access memory (RAM) 903. The processor 901 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 901 may also include onboard memory for caching purposes. The processor 901 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0265] RAM 903 stores various programs and data required for the operation of electronic device 900. Processor 901, ROM 902, and RAM 903 are interconnected via bus 904. Processor 901 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 902 and / or RAM 903. It should be noted that the programs may also be stored in one or more memories other than ROM 902 and RAM 903. Processor 901 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0266] According to embodiments of this application, the electronic device 900 may further include an input / output (I / O) interface 905, which is also connected to a bus 904. The electronic device 900 may also include one or more of the following components connected to the input / output (I / O) interface 905: an input section 906 including a keyboard, mouse, etc.; an output section 907 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the input / output (I / O) interface 905 as needed. A removable medium 911, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 910 as needed so that computer programs read from it can be installed into the storage section 908 as needed.

[0267] According to embodiments of this application, the method flow according to embodiments of this application can be implemented as a computer software program. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 909, and / or installed from removable medium 911. When the computer program is executed by processor 901, it performs the functions defined in the system of embodiments of this application. According to embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0268] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0269] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0270] For example, according to embodiments of this application, a computer-readable storage medium may include the ROM 902 and / or RAM 903 described above and / or one or more memories other than ROM 902 and RAM 903.

[0271] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods provided in the embodiments of this application. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the vibration compensation method provided in the embodiments of this application.

[0272] When the computer program is executed by the processor 901, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0273] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 909, and / or installed from a removable medium 911. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0274] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0275] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations are not explicitly described in this application. In particular, without departing from the spirit and teachings of this application, the features described in the various embodiments of this application can be combined and / or combined in various ways. All such combinations and / or combinations fall within the scope of this application.

[0276] The embodiments of this application have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this application. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Without departing from the scope of this application, those skilled in the art can make various substitutions and modifications, all of which should fall within the scope of this application.

Claims

1. A vibration compensation method, characterized in that, The method includes: Based on the pre-constructed relationship of the vehicle body's pose changing over time under vibration, the vehicle body vibration pose at the time point to be measured is determined. The pose change relationship is constructed based on multiple sets of first offset data and multiple sets of second offset data collected by data acquisition equipment installed on the vehicle body during train operation. The multiple sets of first offset data and multiple sets of second offset data are respectively: multiple sets of offset data on the relative position between the contact rail and the running rail of the train changing over time under the influence of vehicle body vibration and data acquisition equipment vibration; and multiple sets of offset data on the relative position between the vehicle body and the running rail changing over time. Based on the vehicle body vibration posture, determine the motion displacement of the vehicle body relative to the running track at the time point to be measured; The motion displacement is converted into vehicle vibration compensation parameters; Vibration compensation for the train is performed based on the aforementioned vehicle body vibration compensation parameters.

2. The method according to claim 1, characterized in that, The data acquisition equipment includes: a first laser, a first camera, a second laser, and a second camera; The first offset data is collected in the following manner: The first image of the contact rail illuminated by the first laser is captured using the first camera; Based on the line laser on the contact rail in the first image, determine the first offset data of the contact rail relative to the running rail in the cross section perpendicular to the extension direction of the rail in the world coordinate system. The second offset data is collected in the following manner: Multiple second images of the travel track illuminated by the second laser are acquired using the second camera; Based on the pixel coordinates of the line lasers on the travel track in multiple second images, the laser positions of the line lasers corresponding to the multiple second images in the world coordinate system are determined respectively. Second offset data is used to determine the relative position between the vehicle body and the running track based on the laser positions corresponding to multiple second images.

3. The method according to claim 2, characterized in that, The determination of the first offset data of the contact rail relative to the running rail in the world coordinate system on a cross section perpendicular to the rail extension direction, based on the line laser on the contact rail in the first image, includes: The first image is subjected to distortion correction processing to obtain a corrected image; Using the line laser on the contact rail in the first image, determine the key point pixel coordinates of key points on the contact rail in the corrected image; Based on the pixel coordinates of the key points, the three-dimensional spatial offset between the standard positions of the contact rails in the camera coordinate system is determined. Based on the three-dimensional spatial offset, the geometric parameters of the contact rail, and the depth information of the corrected image, the first offset data of the contact rail relative to the running rail on the cross section perpendicular to the track extension direction in the world coordinate system is determined.

4. The method according to claim 3, characterized in that, The step of performing distortion correction processing on the first image to obtain a corrected image includes: The first image is preprocessed to obtain a preprocessed image; Edge extraction is performed on the preprocessed image to obtain the edge bounding box of the laser on the contact rail in the preprocessed image; The preprocessed image is subjected to N distortion correction processes to obtain a corrected image, where N is an integer greater than or equal to 1, and the nth distortion correction process includes: Based on the pixel coordinates of the vertices of the nth bounding box, the preprocessed image is deformed to obtain a deformed image, where n = 1, 2, ..., N; Determine the rectangularity of the bounding box in the deformed image, wherein the rectangularity characterizes how close the shape of the bounding box is to a rectangle; If the rectangularity characterization of the bounding box in the deformed image meets the rectangular requirement, the nth distortion correction process is completed, and the deformed image is determined as the corrected image; If the rectangularity characterization of the bounding box in the deformed image does not meet the rectangularity requirement, the bounding box of the nth time in the deformed image is refitted to obtain the bounding box of the (n+1)th time, and the distortion correction process is performed for the (n+1)th time.

5. The method according to claim 4, characterized in that, The preprocessed image is deformed based on the pixel coordinates of the vertices of the nth bounding box to obtain a deformed image, including: Based on the pixel coordinates of the vertices of the nth bounding box, determine the camera tilt feature when capturing the first image. Based on the tilt feature, the preprocessed image is subjected to a first deformation correction to obtain a preliminary corrected image; Based on the distortion features and intrinsic parameters of the camera that acquired the first image, a second deformation correction is performed on the preliminarily corrected image to achieve deformation removal and obtain a deformed image.

6. The method according to claim 4, characterized in that, The preprocessing of the first image to obtain a preprocessed image includes: The first image is downsampled and filtered to obtain the lower layer image; The lower-layer image is interpolated and dilated to obtain a dilated image; Based on the difference in pixel values ​​between the lower layer image and the dilated image, a multi-detail layer image is determined; The multi-detail layer images are fused to preserve the detail information in each detail layer, resulting in a detail fusion result. Based on the detail fusion results and the multi-detail layer image, image reconstruction is performed to obtain the reconstructed image; The reconstructed image is subjected to adaptive illumination compensation processing to obtain an illumination-compensated image; Connected regions in the illumination-compensated image are marked to obtain a preliminary processed image.

7. The method according to claim 6, characterized in that, The step of performing adaptive illumination compensation processing on the reconstructed image to obtain an illumination-compensated image includes: Background extraction is performed on the reconstructed image to obtain the target object image after removing the background from the reconstructed image; Perform M illumination compensation processes, where M is an integer greater than or equal to 1. The m-th illumination compensation process includes: Determine a first non-uniformity coefficient that characterizes the degree of non-uniformity of grayscale values ​​in the target object image; Based on the m-th illumination compensation coefficient characterizing the illumination supplementation intensity, illumination compensation is performed on the target object image to obtain a preliminary illumination supplementation image, where m = 1, 2, ..., M; Determine a second non-uniformity coefficient that characterizes the degree of gray-level non-uniformity in the preliminary optically compensated image; If the preliminary light-compensated image meets the illumination compensation requirements based on the first non-uniformity coefficient and the second non-uniformity coefficient, then the m-th illumination compensation process is completed, and the preliminary light-compensated image is determined as the illumination compensation image. If the preliminary light-compensated image does not meet the illumination compensation requirements based on the first and second non-uniformity coefficients, then the illumination compensation coefficient for the (m+1)th time is determined, and the (m+1)th compensation process is performed.

8. The method according to claim 3, characterized in that, The step of using the line laser on the contact rail in the first image to determine the key point pixel coordinates of key points on the contact rail in the correction image includes: extracting the contact rail region in the correction image to obtain a contact rail image; Based on the pixel coordinates of the laser lines in the contact rail image, the three-dimensional model of the contact rail is reconstructed to obtain the three-dimensional model of the contact rail in the world coordinate system. Based on the three-dimensional coordinates of the contact rail key points where the calibration area intersects with the laser line in the three-dimensional model of the contact rail, and the correspondence between the pre-calibrated three-dimensional coordinates and sub-pixel coordinates, the sub-pixel level contact rail pixel coordinates of the contact rail key points in the corrected image are determined.

9. The method according to any one of claims 1-8, characterized in that, Also includes: Based on the first offset number, the second offset data, and the error law characterizing the actual position of the contact rail and the observed position of the contact rail, the measurement error caused by vehicle vibration to the position parameter detection of the contact rail is determined. The actual position of the contact rail is the position of the contact rail when only affected by vehicle vibration, and the observed position of the contact rail is the position of the contact rail acquired by the data acquisition device when both the vehicle body and the data acquisition device are affected by vehicle vibration. Based on multiple sets of first offset data and the measurement errors corresponding to multiple sets of first offset data, the vehicle body vibration poses caused by vibration are determined when the vibration causes measurement errors in the data acquisition equipment. The relationship between pose changes is constructed based on multiple sets of vehicle body vibration poses.

10. The method according to claim 9, characterized in that, The determination of multiple sets of vehicle body vibration poses under the condition of measurement error, based on multiple sets of first offset data and the corresponding measurement errors of multiple sets of first offset data, includes: Multiple sets of first offset data and the measurement error corresponding to multiple sets of first offset data are modeled as random variables. The random variables represent the uncertainty of the first offset data collected under the conditions of vehicle body vibration and data acquisition equipment vibration. The pose representation of the random variables corresponding to multiple sets of first offset data is obtained as multiple sets of vehicle body vibration poses generated by the vehicle body under the condition of measurement error.

11. The method according to any one of claims 1-8, characterized in that, The determination of the vehicle body's displacement relative to the running track at the time point to be measured, based on the vehicle body's vibration pose, includes: Bandpass filtering is applied to the vibration pose of the vehicle body to obtain the low-frequency and high-frequency components of the vibration. Based on the low-frequency and high-frequency components of the vibration, the desired pose is converted into an equivalent excitation quantity characterizing the vibration intensity experienced by the vehicle body. Based on the pre-constructed force-displacement relationship between the equivalent excitation and the motion displacement, the motion displacement of the vehicle body relative to the running track at the time point to be measured corresponding to the equivalent excitation is determined. The force-displacement relationship is constructed based on multiple historically acquired motion displacements and multiple equivalent excitations corresponding to the multiple motion displacements.

12. The method according to claim 11, characterized in that, The force-displacement relationship is constructed based on multiple historically acquired motion displacements and their corresponding equivalent excitation quantities in the following manner: Based on multiple motion displacements and the multiple equivalent excitation quantities corresponding to these motion displacements, the correlation between the equivalent excitation quantities and the motion displacements is extracted. Based on the aforementioned correlation pattern, the influence weight of the equivalent excitation on the motion displacement is determined; Perform Z relation construction processes, where Z is an integer greater than or equal to 1. The z-th relation construction process includes: Based on the equivalent incentive amount, the correlation law, and the influence weight of the z-th time, a candidate correspondence is constructed with the equivalent incentive amount as the independent variable and the motion displacement as the dependent variable, where z = 1, 2, ... Z; Candidate motion displacements at test time points are determined based on the candidate correspondence; If the prediction error rate of the candidate motion displacement determined based on the pre-built error prediction model is higher than the preset error threshold, then the adjustment influence weight of the (z+1)th step is determined based on the correlation law, and the (z+1)th step relationship construction processing is performed. If the prediction error rate is less than or equal to a preset error threshold, then the z-th relationship construction process is completed, and the candidate correspondence is determined as a force-displacement relationship.

13. The method according to any one of claims 1-8, characterized in that, The process of converting the motion displacement into vehicle vibration compensation parameters includes: Establish the mapping relationship between the force and displacement transmission characteristics of the vehicle body active suspension system, so as to construct the parameter conversion relationship between motion displacement and vehicle body vibration compensation parameters; Based on the parameter conversion relationship, the motion displacement is converted into vehicle vibration compensation parameters.

14. A vibration compensation device, characterized in that, include: The vehicle body vibration pose determination module is used to determine the vehicle body vibration pose at the time point to be measured by vibration based on the pre-constructed pose change relationship of the vehicle body under vibration influence over time. The pose change relationship is constructed based on multiple sets of first offset data and multiple sets of second offset data collected by data acquisition equipment installed on the vehicle body during train operation. The multiple sets of first offset data and multiple sets of second offset data are respectively: multiple sets of offset data on the relative position between the contact rail and the running rail of the train changing over time under the influence of vehicle body vibration and data acquisition equipment vibration, and multiple sets of offset data on the relative position between the vehicle body and the running rail changing over time. The motion displacement determination module is used to determine the motion displacement of the vehicle body relative to the running track at the time point to be measured, based on the vehicle body vibration posture. The compensation parameter conversion module is used to convert the motion displacement into vehicle vibration compensation parameters; The vibration compensation module is used to perform vibration compensation on the train based on the vehicle body vibration compensation parameters.

15. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs. Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1 to 13.

16. A train, characterized in that, include: The electronic device according to claim 15.

17. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores executable instructions that, when executed by a processor, cause the processor to perform the method of any one of claims 1 to 13.

18. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-13.

Images