Wheel alignment control system and method for electric transport vehicle

Abstract

The present application discloses a wheel alignment control system and method for an electric transport vehicle. The system includes: a wheel integrated into the electric transport vehicle, a multi-line laser profiler, a high-precision camera, a driver, and a control unit; the control unit is used to start the multi-line laser profiler when the electric transport vehicle is in a parked state to emit multiple laser lines to the surface of the moving wheel; receive three-dimensional point cloud data of the wheel fed back by the multi-line laser profiler; receive a two-dimensional image of the wheel about the moving wheel fed back by the high-precision camera; perform multimodal data fusion analysis based on the three-dimensional point cloud data and the two-dimensional image of the wheel to obtain the current actual angle of the moving wheel; and control the driver to rotate according to the current actual angle to rotate the moving wheel to a preset alignment position. By adopting the embodiments of the present application, high-precision wheel alignment can be achieved, thereby improving the accuracy of wheel alignment.

Description

Technical Field

[0001] The present application relates to the technical field of wheel control, and in particular to a wheel alignment control system and method for an electric transport vehicle. Background Art

[0002] Electric trucks are widely used for transporting and handling goods in modern logistics and warehousing, factory material handling, and port loading and unloading. After an electric truck stops and shuts down, if the previous user hasn't properly aligned the wheels, the next user will have difficulty accurately determining the correct direction of the wheels upon starting. This not only increases operational difficulty but can also pose a safety hazard. Therefore, accurately aligning the wheels after parking is crucial.

[0003] Currently, the focus is on simple automatic wheel alignment functions. For example, some electric transport trucks are equipped with an automatic wheel alignment system based on angle sensors, which automatically adjust the wheels to the center position after parking. However, because the angle sensors lose sensitivity over time, high-precision wheel alignment cannot be achieved, thus reducing the accuracy of wheel alignment. Summary of the Invention

[0004] The embodiments of this application provide a wheel alignment control system for an electric transport truck. To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is provided below. This summary is not intended to be a comprehensive review, identify key or important components, or delineate the scope of protection for these embodiments. Its sole purpose is to present some concepts in a simplified form, serving as a prelude to the detailed description that follows.

[0005] In a first aspect, an embodiment of the present application provides a wheel alignment control system for an electric transport vehicle, the system comprising:

[0006] The wheels, multi-line laser profiler, high-precision camera, drive and control unit are integrated into the electric transport vehicle;

[0007] The wheel includes a wheel disc and a mobile wheel. The top of the wheel disc is fixed on the chassis of the electric transport vehicle. A fixing rod is provided on the connecting piece between the wheel disc and the mobile wheel.

[0008] A multi-line laser profiler and a high-precision camera are mounted on a fixed rod at a preset distance from the wheel surface. A connecting shaft is provided inside the wheel disc, one end of which is connected to the moving wheel and the other end is connected to the driver.

[0009] The control unit is connected to the driver, multi-line laser profiler, and high-precision camera respectively; wherein,

[0010] The control unit is used to start the multi-line laser profiler when the vehicle state of the electric transport truck is the parking state, so as to emit multiple laser lines to the surface of the moving wheel; receive the three-dimensional wheel point cloud data about the moving wheel fed back by the multi-line laser profiler; receive the two-dimensional wheel image about the moving wheel fed back by the high-precision camera; perform multimodal data fusion analysis based on the three-dimensional wheel point cloud data and the two-dimensional wheel image to obtain the current actual angle of the moving wheel; and control the rotation of the driver according to the current actual angle to rotate the moving wheel to a preset correction position point.

[0011] In a second aspect, a wheel alignment control method for an electric transport vehicle is provided, which is applied to a control unit of the electric transport vehicle. The method comprises:

[0012] When the electric transport truck is in a parked state, a multi-line laser profiler is activated to emit a plurality of laser lines to the surface of the moving wheel;

[0013] receiving three-dimensional point cloud data of the moving wheel fed back by a multi-line laser profiler;

[0014] receiving a two-dimensional image of the moving wheel fed back by a high-precision camera;

[0015] Perform multimodal data fusion analysis based on the wheel's 3D point cloud data and 2D wheel image to obtain the current actual angle of the moving wheel;

[0016] According to the current actual angle, the driver is controlled to rotate to rotate the moving wheel to the preset home position point.

[0017] In an embodiment of the present application, a control unit activates a multi-line laser profiler when the vehicle is parked, emitting laser lines onto the wheel surface. The control unit then receives feedback, including 3D point cloud data of the wheel and 2D image data from a high-precision camera. The 3D point cloud data provides precise wheel shape and size information, capturing minute surface details and complex shapes, enabling high-precision measurement. Simultaneously, the 2D image provides the overall appearance and edge features of the wheel. By fusing these two types of data for analysis, the accuracy of the wheel alignment control system can be improved, enabling high-precision wheel alignment and thus enhancing wheel alignment accuracy.

[0018] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0020] Figure 1This is a schematic diagram of the system structure of a wheel alignment control system for an electric transport vehicle provided in an embodiment of the present application;

[0021] Figure 2 This is a schematic structural diagram of a wheel provided in an embodiment of the present application;

[0022] Figure 3 This is a side view of an electric transport vehicle provided in an embodiment of the present application;

[0023] Figure 4 is a side schematic diagram of another electric transport vehicle provided in an embodiment of the present application;

[0024] Figure 5 Schematic diagram of a model architecture of a pre-trained tilt angle quantization model provided in an embodiment of the present application;

[0025] Figure 6 This is a flow chart of a model training method for a tilt angle quantization model provided in an embodiment of the present application;

[0026] Figure 7 This is a flow chart of a wheel alignment control method for an electric transport vehicle provided in an embodiment of the present application;

[0027] Figure 8 This is a schematic structural diagram of a control unit provided in an embodiment of the present application;

[0028] Figure 9 This is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The following description and the drawings sufficiently illustrate specific embodiments of the application to enable those skilled in the art to practice them.

[0030] It should be clear that the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0031] When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Instead, they are merely examples of systems and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0032] In the description of this application, it should be understood that the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances. In addition, in the description of this application, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the previous and subsequent associated objects are in an "or" relationship.

[0033] Currently, wheel alignment mainly focuses on simple automatic alignment functions. For example, some electric transport trucks are equipped with an automatic wheel alignment system based on angle sensors, which can automatically adjust the wheels to the center position after parking.

[0034] The inventors have realized that since the sensitivity of the angle sensor decreases with the increase of usage time, high-precision wheel alignment cannot be achieved, thereby reducing the accuracy of wheel alignment.

[0035] In an embodiment of the present application, a control unit activates a multi-line laser profiler when the vehicle is parked, emitting laser lines onto the wheel surface. The control unit then receives feedback, including 3D point cloud data of the wheel and 2D image data from a high-precision camera. The 3D point cloud data provides precise wheel shape and size information, capturing minute surface details and complex shapes, enabling high-precision measurement. Simultaneously, the 2D image provides the overall appearance and edge features of the wheel. By fusing these two types of data for analysis, the accuracy of the wheel alignment control system can be improved, enabling high-precision wheel alignment and thus enhancing wheel alignment accuracy.

[0036] See Figure 1 , Figure 1 The system structure diagram of a wheel alignment control system for an electric transport vehicle provided in an embodiment of the present application is shown. The system comprises: a wheel integrated into the electric transport vehicle, a multi-line laser profiler 1, a high-precision camera (2), a driver, and a control unit. The driver and control unit are inherent components of the electric transport vehicle and are not described in detail here.

[0037] Among them, for example Figure 2As shown, the wheel includes a wheel disc (4) and a movable wheel (5), the top of the wheel disc (4) is fixed on the chassis of the electric transport vehicle; a fixing rod (3) is provided on the connecting piece between the wheel disc (4) and the movable wheel (5); a multi-line laser profiler 1 and a high-precision camera (2) are installed on the fixing rod (3) and are at a preset distance from the wheel surface; a connecting shaft is provided inside the wheel disc (4), one end of the connecting shaft is connected to the movable wheel (5), and the other end is connected to the driver; wherein, the control unit is respectively connected to the driver, the multi-line laser profiler 1, and the high-precision camera (2) for communication. wherein, the connecting shaft is an inherent component of the electric transport vehicle for driving the wheel to move, and is not described in detail here.

[0038] Among them, an electric transport truck is an electrically driven transport vehicle used to move materials or goods in environments such as factories and warehouses. The moving wheels are connected to the wheel discs through connectors. The moving wheels are in direct contact with the ground and are responsible for the movement of the vehicle. A multi-line laser profiler is a measuring device that can emit multiple laser lines and capture their reflections to obtain three-dimensional point cloud data of an object. A high-precision camera is a camera that can capture high-resolution images and is used to obtain two-dimensional image information of the wheels. The drive is a device that controls the rotation of the wheel and can be an electric motor or other type of actuator. The control unit is the control center of the system, responsible for processing data from the sensor and controlling the action of the drive.

[0039] Among them, the side of the electric transport vehicle is Figure 3 and Figure 4 shown.

[0040] In some embodiments of the present application, a control unit is used to start a multi-line laser profiler to emit multiple laser lines to the surface of the moving wheel when the vehicle state of the electric transport truck is a parking state; receive three-dimensional wheel point cloud data about the moving wheel fed back by the multi-line laser profiler; receive a two-dimensional wheel image about the moving wheel fed back by a high-precision camera; perform multimodal data fusion analysis based on the three-dimensional wheel point cloud data and the two-dimensional wheel image to obtain the current actual angle of the moving wheel; and control the rotation of the driver based on the current actual angle to rotate the moving wheel to a preset correction position point.

[0041] The vehicle status is the vehicle's current operating mode, such as driving or parking. The mobile wheels are the actual wheels of the electric transport truck, which can rotate and move. The 3D point cloud data is 3D information about the wheel surface acquired by a multi-line laser profiler. The 2D image is an image of the wheel acquired by a high-precision camera. The current actual angle is the actual spatial orientation angle of the wheel, which is used to determine whether the wheel needs to be adjusted. The preset correction position is the predetermined position to which the wheel needs to be automatically adjusted, specifically the upright position of the wheel, which is parallel to the vehicle body.

[0042] For example, when an electric transport truck stops, the control unit detects that the vehicle is parked. The control unit activates a multi-line laser profiler, which emits multiple laser lines onto the surface of the moving wheel. The multi-line laser profiler scans the wheel, acquiring 3D point cloud data and feeding it back to the control unit. Simultaneously, a high-precision camera captures the wheel image, acquiring 2D image data. The control unit fuses and analyzes the 3D point cloud data and 2D image data to calculate the current actual angle of the wheel. The analysis results indicate that the current angle of the wheel deviates from the preset correction position. Based on the calculated deviation angle, the control unit sends a command to the driver to rotate the wheel. The driver executes the command and precisely adjusts the wheel angle to return it to the preset position.

[0043] In some embodiments of the present application, a specific process of performing multimodal data fusion analysis based on the three-dimensional point cloud data of the wheel and the two-dimensional image of the wheel to obtain the current actual angle of the mobile wheel includes: using a filtering algorithm to smooth the three-dimensional point cloud data of the wheel to obtain smoothed three-dimensional point cloud data of the wheel; performing feature extraction on the smoothed three-dimensional point cloud data of the wheel to obtain the shape, size, edge position and angle information of the mobile wheel; determining the three-dimensional center position, edge points and angle change points of the mobile wheel based on the shape, size, edge position and angle information of the mobile wheel; and based on the three-dimensional center position, edge points and angle change points, Generate the inclination angle of the mobile wheel; perform grayscale processing, edge detection and background separation on the two-dimensional image of the wheel to extract the wheel contour information, which includes the overall appearance or edge features or marking points; determine the rotation angle of the mobile wheel according to the wheel contour information; determine the comprehensive angle of the mobile wheel in the preset spatial coordinate system based on the inclination angle and the rotation angle, and the two straight lines forming the preset correction position are the horizontal axis and the vertical axis of the preset spatial coordinate system, the forward direction of the electric transport truck coincides with the vertical axis, and the wheel center of the mobile wheel is the origin of the preset spatial coordinate system; use the comprehensive angle as the current actual angle of the mobile wheel.

[0044] Specifically, point cloud data contains the three-dimensional coordinate information of every point on an object's surface. Therefore, the shape and size of an object can be measured by calculating distances between points or fitting geometric models. Point cloud data can be used to identify edge points of an object using edge detection algorithms. For example, point cloud-based curvature calculations or deep learning methods can detect edge features on an object's surface. By fitting planes or calculating geometric features in point cloud data, angular information about the object's surface can be obtained. For example, after fitting planes using point cloud data, angular information can be obtained by calculating the angle between the planes.

[0045] Filtering is a signal processing technique used to remove noise and smooth data curves. 3D point cloud data is data on the wheel's surface shape and dimensions obtained through 3D scanning or measurement. Feature extraction extracts useful information from this data, such as shape, size, edge location, and angle. The 3D center position is the geometric center point of the wheel, located in 3D space. Edge points are prominent points on the wheel's profile, typically located at the edges or corners of the profile. Angle change points are points on the wheel's profile where the angle changes significantly. The tilt angle is the degree of inclination of the wheel relative to the horizontal plane. A 2D image is a two-dimensional image of the wheel, such as a photograph taken by a camera. Grayscaling converts a color image into a grayscale image to simplify data processing. Edge detection is an algorithm that identifies the edges of objects in an image. Background segmentation is a technique for separating the foreground and background from an image. Contour information includes information on the wheel's overall appearance, edge features, and markers. The rotation angle is the angle of rotation of the wheel around its central axis. The combined angle is the final angle of the wheel in the spatial coordinate system after combining the tilt angle and the rotation angle. The preset spatial coordinate system is the reference coordinate system set for the electric transport truck. The current actual angle is the actual spatial direction angle of the wheel and is used for control and adjustment.

[0046] In an embodiment of the present application, by using a filtering algorithm to smooth the three-dimensional point cloud data of the wheel, noise and interference in the data can be effectively removed, thereby obtaining more accurate and reliable three-dimensional point cloud data of the wheel. This step is crucial for subsequent feature extraction because it ensures that the extracted features more accurately reflect the actual shape and size of the wheel. Furthermore, feature extraction based on the smoothed three-dimensional point cloud data can obtain the shape, size, edge position and angle information of the wheel. This information is the basis for determining the three-dimensional center position, edge points and angle change points of the wheel. By determining these key points, the inclination angle of the wheel can be generated, providing important data for subsequent wheel posture analysis.

[0047] In an embodiment of the present application, the two-dimensional image of the wheel is gray-scaled, edge detected, and background separated, and the contour information of the wheel, including the overall appearance, edge features, and markers, can be extracted. This information can help determine the rotation angle of the wheel, further enriching the description of the wheel posture. Ultimately, based on the tilt angle and the rotation angle, the comprehensive angle of the wheel in the preset spatial coordinate system can be determined. The determination of this comprehensive angle enables the current position and posture of the wheel to be accurately described and controlled. Specifically, the two straight lines of the preset return position serve as the horizontal axis and the vertical axis of the coordinate system, the forward direction of the electric transport truck coincides with the vertical axis, and the center of the wheel is the origin of the coordinate system. This setting makes the control and positioning of the wheel more intuitive and accurate. Using the comprehensive angle as the current actual angle of the wheel can provide an accurate reference for the control and adjustment of the wheel, ensuring that the wheel can be accurately rotated to the preset return position, thereby improving the control accuracy and reliability of the entire system.

[0048] In some embodiments of the present application, the specific process of determining the three-dimensional center position, edge point and angle change point of the mobile wheel based on the shape, size, edge position and angle information of the mobile wheel includes: analyzing the shape and size information of the mobile wheel to determine the wheel profile of the mobile wheel; calculating the center of mass of the wheel profile of the mobile wheel, and taking the center of mass as the three-dimensional center position of the mobile wheel in three-dimensional space; calculating the curvature of each contour point on the contour line of the wheel contour through the edge position; taking the contour point with the largest curvature change as the edge point; calculating the angle change of each contour point on the contour line of the wheel contour through the angle information; and taking the contour point with the largest angle change as the angle change point.

[0049] The center of mass of the wheel profile is calculated as:

[0050]

[0051] The contour of the moving wheel is given by points( ) and the center of mass is ( ),( ) is the first contour of the moving wheel The coordinates of the points, is the total number of contour points of the moving wheel; where,

[0052] Edge points ( ) is calculated as:

[0053]

[0054] in, The contour of the moving wheel provided for the edge position The coordinates of the contour points, and For the The first-order derivative of the contour point, that is, the contour at and The slope in the direction, and For the The second-order derivative of the contour point, that is, the contour at and The rate of change of the slope in the direction is used to reflect the profile of the moving wheel in and The curvature changes in the direction;

[0055] Angle change point The calculation formula is:

[0056]

[0057] in, The angle information provided by The angle change of each contour point.

[0058] In some embodiments of the present application, the specific process of generating the tilt angle of the mobile wheel based on the three-dimensional center position, edge points and angle change points includes: standardizing the three-dimensional center position, edge points and angle change points to eliminate the dimensional differences between different features, and obtain the target three-dimensional center position, target edge points and target angle change points; extracting the three-dimensional center position features from the target three-dimensional center position; extracting the edge features from the target edge points; extracting the angle change features from the target angle change points; performing feature splicing processing on the three-dimensional center position features, edge features and angle change features to obtain a comprehensive feature vector that can be understood by a pre-trained tilt angle quantization model; inputting the comprehensive feature vector into the pre-trained tilt angle quantization model to output the tilt angle of the mobile wheel.

[0059] Among them, the three-dimensional center position is the center point of the wheel in three-dimensional space, which is used to determine the position and posture of the wheel. The edge point is the key point on the wheel contour, usually located at the edge or corner of the wheel, and is used to determine the shape and position of the wheel. The angle change point is the point on the wheel contour where the angle changes significantly, which is used to determine the tilt and rotation state of the wheel. Grayscale processing is the process of converting a color image into a grayscale image, which simplifies the image data and facilitates subsequent processing. Edge detection is an algorithm that identifies the edges of objects in an image and is used to extract the contour information of the wheel. Background separation is a technology that separates the foreground and background from the image, which is used to extract the contour information of the wheel. Contour information includes the overall appearance of the wheel, edge features, and marker information, which is used to determine the rotation angle of the wheel.

[0060] Among them, for example Figure 5As shown, the pre-trained tilt angle quantization model includes a feature extraction layer, a feature fusion layer, an angle calculation layer, and an angle output layer.

[0061] In some embodiments of the present application, the comprehensive feature vector is input into a pre-trained tilt angle quantization model, and the specific process of outputting the tilt angle of the mobile wheel includes: the feature extraction layer extracts the feature representations of the three-dimensional center position feature, the edge feature, and the angle change feature from the comprehensive feature vector; the feature fusion layer fuses the feature representation with the preset enhanced feature to form a fused feature vector, and the preset enhanced feature includes the radius feature, the width feature, and the thickness feature of the mobile wheel; the angle calculation layer calculates based on the fused feature to obtain the angle parameter; the angle parameter is decoded and converted into an actual physical angle value to obtain the tilt angle of the mobile wheel; wherein the angle parameter The calculation formula is as follows:

[0062]

[0063] in, is the angle parameter, are the model weights of the pre-trained tilt angle quantization model, is the bias vector of the model, is the fusion feature vector;

[0064]

[0065] in, is the tilt angle of the moving wheel, is the mean value of the tilt angle, is the standard deviation of the tilt angle.

[0066] Specifically, the specific process of generating a pre-trained tilt angle quantization model includes: collecting historical three-dimensional point cloud data of different mobile wheels in a parked state, and based on the historical three-dimensional point cloud data, determining the historical wheel shape, historical size, historical edge position and historical angle information of the mobile wheel; obtaining the historical tilt angles of different mobile wheels in a parked state as tilt angle labels of different mobile wheels; associating the tilt angle labels of different mobile wheels with the historical three-dimensional point cloud data of different mobile wheels in a parked state to obtain model training samples; using a neural network to create a tilt angle quantization model; inputting the model training samples into the pre-trained tilt angle quantization model, and outputting the loss value of the model; when the loss value of the model reaches the minimum, generating a pre-trained tilt angle quantization model.

[0067] Specifically, when the loss value of the model has not reached the minimum, the step of inputting the model training sample into the pre-trained tilt angle quantization model is continued until the loss value of the model reaches the minimum.

[0068] In some embodiments of the present application, the specific process of determining the rotation angle of the mobile wheel based on the wheel profile information includes: when the wheel profile information contains a marker point, extracting the position coordinates of the marker point in the two-dimensional image of the wheel; analyzing the position change of the marker point at different time points based on the position coordinates; or, when the wheel profile information contains edge features, determining the edge position point on the mobile wheel based on the edge features; analyzing the position change of the edge position point at different time points; or, when the wheel profile information contains an overall appearance, extracting the outline of the mobile wheel from the overall appearance; analyzing the position change of the outline at different time points; and calculating the rotation angle of the mobile wheel based on the position change; wherein the calculation formula of the rotation angle is:

[0069]

[0070] in,( )and( ) are the position coordinates of the marking point, edge position point, and contour at different time points.

[0071] The tilt angle is obtained by rotating around the Z axis of the preset spatial coordinate system, and the rotation angle is obtained by rotating around the Y axis of the preset spatial coordinate system.

[0072] In some embodiments of the present application, the specific process of determining the comprehensive angle of the mobile wheel in the preset spatial coordinate system based on the tilt angle and the rotation angle includes: calculating the rotation matrix around the Z axis and the Y axis of the preset spatial coordinate system according to the tilt angle and the rotation angle; wherein,

[0073]

[0074]

[0075] According to the rotation matrix around the Z axis and Y axis of the preset spatial coordinate system, the comprehensive rotation matrix is calculated; wherein the comprehensive rotation matrix The calculation formula is:

[0076]

[0077] in, is the tilt angle around the Z axis, is the rotation angle around the Y axis, and are the rotation matrices around the Z axis and Y axis respectively;

[0078] Based on the comprehensive rotation matrix, calculate the comprehensive angle of the moving wheel in the preset spatial coordinate system ;in,

[0079]

[0080] in, is the trace of the rotation matrix.

[0081] In some embodiments of the present application, the specific process of controlling the rotation of the driver based on the current actual angle to rotate the mobile wheel to a preset home position includes calculating the deviation between the current actual angle and the angle at the preset home position. This deviation angle is used to guide the driver's adjustment actions to ensure that the wheel can accurately rotate to the preset position. Based on the calculated angle deviation, the control unit sends a corresponding control signal to the driver. After receiving the signal, the driver adjusts its output to achieve precise rotation of the wheel. Specifically, this involves the generation of a PWM (pulse width modulation) signal to control the speed and direction of the motor. The driver rotates the wheel from its current position to the preset home position according to the control unit's instructions. During this process, the control unit may also monitor the wheel's position and angle in real time to ensure accurate rotation. Once the wheel rotates to the preset position, the control unit stops the driver's output, completing the entire home position.

[0082] In some embodiments, the connection between the wheel disc and the mobile wheel can be fixed to the frame of the mobile transport vehicle.

[0083] In an embodiment of the present application, a control unit activates a multi-line laser profiler when the vehicle is parked, emitting laser lines onto the wheel surface. The control unit then receives feedback, including 3D point cloud data of the wheel and 2D image data from a high-precision camera. The 3D point cloud data provides precise wheel shape and size information, capturing minute surface details and complex shapes, enabling high-precision measurement. Simultaneously, the 2D image provides the overall appearance and edge features of the wheel. By fusing these two types of data for analysis, the accuracy of the wheel alignment control system can be improved, enabling high-precision wheel alignment and thus enhancing wheel alignment accuracy.

[0084] See Figure 6 , provides a flow chart of a model training method for a tilt angle quantization model according to an embodiment of the present application. Figure 6 As shown, the detection method of the embodiment of the present application may include the following steps:

[0085] S101, collecting historical three-dimensional point cloud data of different mobile wheels in a parked state, and determining historical wheel shape, historical size, historical edge position, and historical angle information of the mobile wheels based on the historical three-dimensional point cloud data;

[0086] S102, obtaining historical tilt angles of different moving wheels in a parked state as tilt angle labels of the different moving wheels;

[0087] S103, associating the tilt angle labels of different moving wheels with the historical three-dimensional point cloud data of the different moving wheels in the parking state to obtain model training samples;

[0088] S104, creating a tilt angle quantization model using a neural network;

[0089] S105, inputting the model training sample into the pre-trained tilt angle quantization model, and outputting the loss value of the model;

[0090] S106: When the loss value of the model reaches a minimum, a pre-trained tilt angle quantization model is generated.

[0091] Wherein, if the loss value of the model has not reached the minimum, the step of inputting the model training sample into the pre-trained tilt angle quantization model is continued until the loss value of the model reaches the minimum.

[0092] It should be noted that the specific control logic of steps S101 to S106 can be found in the specific implementation logic of the system embodiment, and will not be repeated here.

[0093] In an embodiment of the present application, a control unit activates a multi-line laser profiler when the vehicle is parked, emitting laser lines onto the wheel surface. The control unit then receives feedback, including 3D point cloud data of the wheel and 2D image data from a high-precision camera. The 3D point cloud data provides precise wheel shape and size information, capturing minute surface details and complex shapes, enabling high-precision measurement. Simultaneously, the 2D image provides the overall appearance and edge features of the wheel. By fusing these two types of data for analysis, the accuracy of the wheel alignment control system can be improved, enabling high-precision wheel alignment and thus enhancing wheel alignment accuracy.

[0094] See Figure 7 , is a flow chart of a wheel alignment control method for an electric transport vehicle according to an embodiment of the present application. Figure 7 As shown, the detection method of the embodiment of the present application may include the following steps:

[0095] S201, when the vehicle state of the electric transport vehicle is a parking state, starting a multi-line laser profiler to emit multiple laser lines to the surface of the moving wheel;

[0096] S202, receiving three-dimensional point cloud data of the moving wheel fed back by a multi-line laser profiler;

[0097] S203, receiving a two-dimensional image of the moving wheel fed back by a high-precision camera;

[0098] S204, performing multimodal data fusion analysis based on the three-dimensional point cloud data of the wheel and the two-dimensional image of the wheel to obtain the current actual angle of the moving wheel;

[0099] S205 , controlling the driver to rotate according to the current actual angle to rotate the moving wheel to a preset home position point.

[0100] In an embodiment of the present application, a control unit activates a multi-line laser profiler when the vehicle is parked, emitting laser lines onto the wheel surface. The control unit then receives feedback, including 3D point cloud data of the wheel and 2D image data from a high-precision camera. The 3D point cloud data provides precise wheel shape and size information, capturing minute surface details and complex shapes, enabling high-precision measurement. Simultaneously, the 2D image provides the overall appearance and edge features of the wheel. By fusing these two types of data for analysis, the accuracy of the wheel alignment control system can be improved, enabling high-precision wheel alignment and thus enhancing wheel alignment accuracy.

[0101] The following are device embodiments of the present application, which can be used to implement the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

[0102] See Figure 8 , which shows a schematic diagram of the structure of a control unit provided by an exemplary embodiment of the present application. The control unit can be implemented as all or part of an electronic device through software, hardware, or a combination of both. The control unit 1 includes a startup module 10, a first receiving module 20, a second receiving module 30, a fusion analysis module 40, and a control module 50.

[0103] The starting module 10 is used to start the multi-line laser profiler when the vehicle state of the electric transport vehicle is a parking state, so as to emit multiple laser lines to the surface of the moving wheel;

[0104] The first receiving module 20 is configured to receive three-dimensional point cloud data of the moving wheel fed back by the multi-line laser profiler;

[0105] The second receiving module 30 is configured to receive a two-dimensional image of the moving wheel fed back by a high-precision camera;

[0106] A fusion analysis module 40 is used to perform multimodal data fusion analysis based on the three-dimensional point cloud data of the wheel and the two-dimensional image of the wheel to obtain the current actual angle of the moving wheel;

[0107] The control module 50 is used to control the driver to rotate according to the current actual angle, so as to rotate the moving wheel to a preset home position point.

[0108] It should be noted that the wheel alignment control device for an electric truck provided in the above embodiment, when executing the wheel alignment control method for an electric truck, only uses the division of the above-mentioned functional modules as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the wheel alignment control device for an electric truck provided in the above embodiment and the wheel alignment control method embodiment for an electric truck are based on the same concept. The implementation process is detailed in the method embodiment and will not be repeated here.

[0109] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0110] In an embodiment of the present application, a control unit activates a multi-line laser profiler when the vehicle is parked, emitting laser lines onto the wheel surface. The control unit then receives feedback, including 3D point cloud data of the wheel and 2D image data from a high-precision camera. The 3D point cloud data provides precise wheel shape and size information, capturing minute surface details and complex shapes, enabling high-precision measurement. Simultaneously, the 2D image provides the overall appearance and edge features of the wheel. By fusing these two types of data for analysis, the accuracy of the wheel alignment control system can be improved, enabling high-precision wheel alignment and thus enhancing wheel alignment accuracy.

[0111] The present application also provides a computer-readable medium having program instructions stored thereon, which, when executed by a processor, implement the wheel alignment control method for the electric transport vehicle provided by the above-mentioned various method embodiments.

[0112] The present application also provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the wheel alignment control method for an electric transport vehicle according to each of the above method embodiments.

[0113] See Figure 9 , is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 9 As shown, the electronic device 1000 may include: at least one processor 1001 , at least one network interface 1004 , a user interface 1003 , a memory 1005 , and at least one communication bus 1002 .

[0114] The communication bus 1002 is used to implement the connection and communication between these components.

[0115] The user interface 1003 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 1003 may also include a standard wired interface and a wireless interface.

[0116] The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0117] The processor 1001 may include one or more processing cores. The processor 1001 utilizes various interfaces and circuits to connect various components within the electronic device 1000. It executes instructions, programs, code sets, or instruction sets stored in the memory 1005, and accesses data stored in the memory 1005 to perform various functions and process data within the electronic device 1000. Optionally, the processor 1001 may be implemented in hardware using at least one of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 1001 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing content displayed on the display; and the modem handles wireless communications. It is understood that the modem may also be implemented independently of the processor 1001 and implemented on a separate chip.

[0118] Among them, the memory 1005 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory 1005 includes a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 1005 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 1005 may also be optionally at least one storage system located away from the aforementioned processor 1001. As Figure 9As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a wheel alignment control application for the electric transport vehicle.

[0119] exist Figure 9 In the electronic device 1000 shown, the user interface 1003 is mainly used to provide an input interface for the user and obtain user input data; and the processor 1001 can be used to call the wheel alignment control application for the electric transport vehicle stored in the memory 1005 and specifically perform the following operations:

[0120] When the electric transport truck is in a parked state, a multi-line laser profiler is activated to emit a plurality of laser lines to the surface of the moving wheel;

[0121] receiving three-dimensional point cloud data of the moving wheel fed back by a multi-line laser profiler;

[0122] receiving a two-dimensional image of the moving wheel fed back by a high-precision camera;

[0123] Perform multimodal data fusion analysis based on the wheel's 3D point cloud data and 2D wheel image to obtain the current actual angle of the moving wheel;

[0124] According to the current actual angle, the driver is controlled to rotate to rotate the moving wheel to the preset home position point.

[0125] In an embodiment of the present application, a control unit activates a multi-line laser profiler when the vehicle is parked, emitting laser lines onto the wheel surface. The control unit then receives feedback, including 3D point cloud data of the wheel and 2D image data from a high-precision camera. The 3D point cloud data provides precise wheel shape and size information, capturing minute surface details and complex shapes, enabling high-precision measurement. Simultaneously, the 2D image provides the overall appearance and edge features of the wheel. By fusing these two types of data for analysis, the accuracy of the wheel alignment control system can be improved, enabling high-precision wheel alignment and thus enhancing wheel alignment accuracy.

[0126] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program for controlling the wheel alignment of an electric transport vehicle can be stored in a computer-readable storage medium. When executed, the program can include the processes in the above-described method embodiments. The storage medium for the program for controlling the wheel alignment of an electric transport vehicle can be a magnetic disk, an optical disk, a read-only memory, or a random access memory.

[0127] The above disclosure is only a preferred embodiment of the present application, and certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A wheel alignment control system for an electric transport vehicle, characterized in that: The system comprises: The wheels, multi-line laser profiler, high-precision camera, driver and control unit are integrated into the electric transport vehicle; wherein, The wheel includes a wheel disc and a movable wheel, the top of the wheel disc is fixed to the chassis of the electric transport vehicle; a fixing rod is provided on the connecting piece between the wheel disc and the movable wheel; The multi-line laser profiler and the high-precision camera are mounted on the fixed rod and at a preset distance from the wheel surface. A connecting shaft is provided inside the wheel disc, one end of which is connected to the moving wheel and the other end is connected to the driver. The control unit is respectively connected to the driver, the multi-line laser profiler, and the high-precision camera for communication; wherein, The control unit is configured to, when the vehicle state of the electric transport truck is a parking state, start the multi-line laser profiler to emit multiple laser lines to the surface of the mobile wheel; receive the three-dimensional point cloud data of the mobile wheel fed back by the multi-line laser profiler, the three-dimensional point cloud data being three-dimensional information of the wheel surface; receive the two-dimensional wheel image of the mobile wheel fed back by the high-precision camera; perform multimodal data fusion analysis based on the three-dimensional point cloud data of the wheel and the two-dimensional wheel image, including: performing feature extraction on the smoothed three-dimensional point cloud data of the wheel to obtain the shape, size, edge position and angle information of the mobile wheel; determining the three-dimensional center position, edge point and angle change point of the mobile wheel based on the shape, size, edge position and angle information of the mobile wheel; generating the inclination angle of the mobile wheel based on the three-dimensional center position, edge point and angle change point; extracting wheel contour information from the two-dimensional wheel image, and determining the rotation angle of the mobile wheel based on the wheel contour information; The tilt angle is obtained by rotating around the Z axis of the preset spatial coordinate system, and the rotation angle is obtained by rotating around the Y axis of the preset spatial coordinate system; based on the tilt angle and the rotation angle, a rotation matrix around the Z axis and the Y axis of the preset spatial coordinate system is calculated; based on the rotation matrix around the Z axis and the Y axis of the preset spatial coordinate system, a comprehensive rotation matrix is calculated; based on the comprehensive rotation matrix, a comprehensive angle of the mobile wheel in the preset spatial coordinate system is calculated; and the comprehensive angle is used as the current actual angle of the mobile wheel; The current actual angle of the moving wheel is obtained; and according to the current actual angle, the driver is controlled to rotate so as to rotate the moving wheel to a preset home position point.

2. The system according to claim 1, wherein: The performing multimodal data fusion analysis based on the three-dimensional point cloud data of the wheel and the two-dimensional image of the wheel to obtain the current actual angle of the moving wheel includes: Smoothing the wheel three-dimensional point cloud data using a filtering algorithm to obtain smoothed wheel three-dimensional point cloud data; grayscale processing, edge detection, and background separation are performed on the two-dimensional wheel image to extract wheel profile information, wherein the wheel profile information includes overall appearance or edge features or marking points; Based on the tilt angle and the rotation angle, the comprehensive angle of the mobile wheel in the preset spatial coordinate system is determined. The two straight lines forming the preset correction position are the horizontal axis and the vertical axis of the preset spatial coordinate system, respectively. The forward direction of the electric transport vehicle coincides with the longitudinal axis, and the wheel center of the mobile wheel is the origin of the preset spatial coordinate system.

3. The system according to claim 2, characterized in that The determining of the three-dimensional center position, edge point, and angle change point of the mobile wheel according to the shape, size, edge position, and angle information of the mobile wheel comprises: analyzing the shape and size information of the mobile wheel to determine the wheel profile of the mobile wheel; Calculating the center of mass of the wheel profile of the mobile wheel, and using the center of mass as the three-dimensional center position of the mobile wheel in the three-dimensional space; Calculating the curvature of each contour point on the contour line of the wheel contour according to the edge position; The contour point with the largest curvature change is regarded as the edge point; Calculating the angle change of each contour point on the contour line of the wheel contour using the angle information; The contour point with the largest angle change is taken as the angle change point.

4. The system according to claim 2, wherein: Generating the tilt angle of the moving wheel based on the three-dimensional center position, the edge point, and the angle change point includes: Normalizing the three-dimensional center position, edge point, and angle change point to eliminate dimensional differences between different features, thereby obtaining a target three-dimensional center position, target edge point, and target angle change point; Extracting a three-dimensional center position feature from the target three-dimensional center position; Extracting edge features from the target edge points; Extracting angle change features from the target angle change points; Performing feature splicing processing on the three-dimensional center position feature, the edge feature, and the angle change feature to obtain a comprehensive feature vector that can be understood by a pre-trained tilt angle quantization model; The comprehensive feature vector is input into a pre-trained tilt angle quantization model to output the tilt angle of the moving wheel.

5. The system according to claim 4, characterized in that The pre-trained tilt angle quantization model includes a feature extraction layer, a feature fusion layer, an angle calculation layer and an angle output layer; Inputting the comprehensive feature vector into a pre-trained tilt angle quantization model to output the tilt angle of the moving wheel includes: The feature extraction layer extracts feature representations of three-dimensional center position features, edge features, and angle change features from the comprehensive feature vector; The feature fusion layer fuses the feature representation with preset enhanced features to form a fused feature vector, wherein the preset enhanced features include a radius feature, a width feature, and a thickness feature of the moving wheel; The angle calculation layer performs calculation based on the fusion features to obtain angle parameters; The angle parameter is decoded and converted into an actual physical angle value to obtain the inclination angle of the moving wheel.

6. The system according to claim 4, characterized in that Follow these steps to generate a pre-trained tilt angle quantization model, including: collecting historical three-dimensional point cloud data of different mobile wheels in a parked state, and determining historical wheel shapes, historical dimensions, historical edge positions, and historical angle information of the mobile wheels based on the historical three-dimensional point cloud data; Obtain historical tilt angles of different moving wheels in a parked state as tilt angle labels of different moving wheels; The tilt angle labels of different moving wheels are associated with the historical 3D point cloud data of different moving wheels in the parking state to obtain model training samples; A neural network is used to create a tilt angle quantification model; Inputting the model training sample into the pre-trained tilt angle quantization model, and outputting the loss value of the model; When the loss value of the model reaches a minimum, a pre-trained tilt angle quantization model is generated.

7. The system according to claim 6, characterized in that The system further comprises: In the case that the loss value of the model has not reached the minimum, the step of inputting the model training samples into the pre-trained tilt angle quantization model is continued until the loss value of the model reaches the minimum.

8. The system according to claim 2, wherein: Determining the rotation angle of the moving wheel according to the wheel profile information includes: If the marking point exists in the wheel profile information, extracting the position coordinates of the marking point in the two-dimensional image of the wheel; Analyzing the position changes of the marked points at different time points using the position coordinates; or, In a case where the wheel profile information includes the edge feature, determining an edge position point on the moving wheel based on the edge feature; Analyzing the position changes of the edge position points at different time points; or, extracting the contour of the moving wheel from the overall appearance when the wheel contour information has the overall appearance; analyzing position changes of the contour at different time points; The rotation angle of the moving wheel is calculated according to the position change.

9. A wheel alignment control method for an electric transport vehicle implemented using the system according to any one of claims 1 to 8, characterized in that: Applied to the control unit of the electric transport vehicle, the method comprises: When the electric transport vehicle is in a parked state, starting the multi-line laser profiler to emit a plurality of laser lines to the surface of the moving wheel; receiving three-dimensional point cloud data of the moving wheel fed back by the multi-line laser profiler; receiving a two-dimensional wheel image of the moving wheel fed back by the high-precision camera; Performing multimodal data fusion analysis on the three-dimensional point cloud data of the wheel and the two-dimensional image of the wheel to obtain the current actual angle of the mobile wheel; comprising: performing feature extraction on the smoothed three-dimensional point cloud data of the wheel to obtain the shape, size, edge position and angle information of the mobile wheel; determining the three-dimensional center position, edge points and angle change points of the mobile wheel based on the shape, size, edge position and angle information of the mobile wheel; generating the inclination angle of the mobile wheel based on the three-dimensional center position, edge points and angle change points; extracting wheel contour information from the two-dimensional image of the wheel, and determining the rotation angle of the mobile wheel based on the wheel contour information; The tilt angle is obtained by rotating around the Z axis of the preset spatial coordinate system, and the rotation angle is obtained by rotating around the Y axis of the preset spatial coordinate system; based on the tilt angle and the rotation angle, a rotation matrix around the Z axis and the Y axis of the preset spatial coordinate system is calculated; based on the rotation matrix around the Z axis and the Y axis of the preset spatial coordinate system, a comprehensive rotation matrix is calculated; based on the comprehensive rotation matrix, a comprehensive angle of the mobile wheel in the preset spatial coordinate system is calculated; and the comprehensive angle is used as the current actual angle of the mobile wheel; According to the current actual angle, the driver is controlled to rotate so as to rotate the moving wheel to a preset home position point.

Images