Vehicle automatic alignment method, electronic system, medium and program product

Abstract

The invention discloses an automatic vehicle alignment method, an electronic system, a medium and a program product, which can quickly improve the control precision of automatic vehicle alignment at low cost. The automatic vehicle alignment method is a method for automatically aligning a trailer of a trailer to a conveyor belt, and comprises the following steps: identifying one of a front edge and a rear edge of at least one trailer of the trailer in a driving direction of the trailer; calculating an actual position coordinate of one of the front edge and the rear edge according to an identification result; according to the actual position coordinates and preset stopping position coordinates, the offset of the trailer relative to the conveying belt is determined; and moving the trailer relative to the conveyor belt to align with the conveyor belt according to the offset.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle automatic alignment and vehicle management, and more particularly to a vehicle automatic alignment method, an electronic system, a readable storage medium, and a computer program product. Background Art

[0002] It is known that in docks and airport warehouses, it is necessary to operate trailers in a narrow space to accurately align them with the conveyor belt before starting to convey goods. If there is a misalignment between the trailer and the conveyor belt during the alignment process, it may cause equipment damage or personal injury. Therefore, alignment not only requires accuracy but also high safety.

[0003] In the past, manual alignment was mostly used. This highly depends on the experience and skills of the operator, often requiring very high driving skills. Novice or inexperienced operators may find it difficult to quickly and accurately complete the alignment operation. In addition, if the on-site space is insufficient or the environment is complex (such as obstacles, other vehicles), the operator's vision may be restricted, making alignment more difficult, increasing the time cost, and having higher technical requirements. Especially in the case of poor visibility or severe occlusion, it is necessary to rely on the monitoring system or auxiliary equipment, which undoubtedly increases the complexity and relative cost of the operation.

[0004] Therefore, an automatic alignment system has emerged. Most automatic alignment systems on the market currently use sensors (such as cameras, etc.) to sense the position and angle of the trailer and the conveyor belt, so as to achieve alignment. However, this not only requires real-time processing of a large amount of sensor data but also has high requirements for the accuracy of the sensors. For example, in the case of uneven terrain, obstacles, and / or large tail swing of the vehicle, the system may not be able to accurately calculate and execute the alignment operation. Summary of the Invention

[0005] Technical Problems to be Solved by the Invention

[0006] The present application is formed to solve the above technical problems, and its purpose is to provide a vehicle automatic alignment method that can accurately detect the relative position relationship and posture between the trailer and the conveyor belt through a specific algorithm, and achieve automatic alignment of the vehicle-conveyor belt with low cost, full automation, high alignment accuracy, high control precision, and high vehicle adaptability. On this basis, the present application further provides an electronic system, a readable storage medium, and a computer program product corresponding to the above control method.

[0007] Technical Solutions for Solving the Technical Problems

[0008] The present application provides a vehicle automatic alignment method, which is a method for automatically aligning the trailer of a trailer with a conveyor belt, including:

[0009] Identify at least one of the front edge and the rear edge of the trailer hitch of the trailer in the traveling direction of the trailer;

[0010] Calculate the actual position coordinates of at least one of the front edge and the rear edge according to the recognition result;

[0011] Determine the offset of the trailer hitch relative to the conveyor belt according to the actual position coordinates and the preset docking position coordinates; and

[0012] Move the trailer hitch relative to the conveyor belt according to the offset to align it with the conveyor belt.

[0013] Preferably, identifying at least one of the front edge and the rear edge of the trailer hitch of the trailer in the traveling direction of the trailer hitch includes:

[0014] Scan the target scene including the at least one trailer hitch by lidar to form a point cloud of the target scene;

[0015] Perform target recognition and extraction on the point cloud of the target scene to identify and extract a point cloud set constituting the at least one trailer hitch from the point cloud;

[0016] Divide the point cloud set of the at least one trailer hitch into multiple point cloud subsets of the same shape along the traveling direction of the trailer; and

[0017] Identify the point cloud subset constituting at least one of the front edge or the rear edge by comparing the number of points in adjacent point cloud subsets.

[0018] Preferably, when the lidar scans, parse the received raw point cloud data to obtain point cloud data that converts angle and distance information into three-dimensional coordinate XYZ values, and adjust the extrinsic parameters of the point cloud data based on the installation angle and installation height of the lidar to align the lidar coordinate system and the world coordinate system.

[0019] Preferably, when performing target recognition and extraction on the point cloud of the target scene, remove the data outside the height range based on the pre-input height range to form the point cloud set.

[0020] Preferably, when dividing the point cloud set into multiple point cloud subsets, slice the point cloud set into multiple thin sheet-shaped point cloud subsets arranged along the traveling direction of the trailer hitch.

[0021] Preferably, the point cloud subset is in the shape of a thin sheet of a cuboid with a length of 0.5 - 3m × a height of 0.1 - 1m × a width of 0.01 - 0.1m.

[0022] Preferably, when the difference between the number of points in one point cloud subset and the number of points in another point cloud subset that are adjacent and arranged in the multiple point cloud subsets reaches a preset first threshold, it is determined that one of the two point cloud subsets is the point cloud subset constituting the front edge or the rear edge.

[0023] Preferably, if the number of points in the previous point cloud subset arranged along the driving direction of the trailer is more than the number of points in the adjacent subsequent point cloud subset and exceeds the first threshold, it is determined as the rear edge of the trailer hitch;

[0024] If the number of points in the subsequent point cloud subset arranged along the driving direction of the trailer is more than the number of points in the adjacent previous point cloud subset and exceeds the first threshold, it is determined as the rear edge of the trailer hitch.

[0025] Preferably, calculating the actual position coordinates of one of the front edge and the rear edge according to the recognition result includes:

[0026] Fitting the plane where one of the front edge or the rear edge is located according to the points in the point cloud subset constituting one of the front edge or the rear edge; and

[0027] Determining the actual position coordinates of one of the front edge or the rear edge according to the coordinate information of the fitted plane.

[0028] Preferably, the coordinate information of the plane of the front edge or the rear edge is fitted by using the RANSAC algorithm.

[0029] Preferably, between two adjacent conveyor belts in the driving direction of the trailer, a lidar is arranged at a specified height from the ground and on the side close to the trailer hitch;

[0030] The laser scanning range of the lidar is at least greater than the interval between adjacent trailer hitches in the driving direction of the trailer.

[0031] Preferably, the trailer travels on a flat ground in one of a dock, an airport cargo warehouse, and a logistics warehouse.

[0032] Advantages of the Invention

[0033] According to the vehicle automatic alignment method described in the present application, compared with the existing vehicle automatic alignment methods, it can achieve the automatic alignment of the trailer hitch of the trailer and the conveyor belt through accurate edge recognition and position calculation, improve the operation efficiency and accuracy, and improve the control accuracy of vehicle automatic alignment at a lower cost. Brief Description of the Drawings

[0034] Figure 1It is a flowchart showing a vehicle automatic alignment method according to an embodiment of the present application.

[0035] Figure 2 It is a schematic structural diagram showing an embodiment of the present application. Detailed implementation manners

[0036] Hereinafter, the present application will be further described in conjunction with the following embodiments. It should be understood that the following embodiments are only used to illustrate the present application and do not limit the present application. Before specific description, for the convenience of clear description and understanding, here, some technical terms and / or technical terms involved in the present application are described or defined.

[0037] Refer to Figure 1 to describe the implementation steps of the vehicle automatic alignment method according to an embodiment of the present application. It should be noted that the steps shown in the following embodiments are only examples. As long as the same technical purpose or a further technical purpose can be achieved, the order of these steps can be appropriately changed, and additional other steps can also be added.

[0038] The vehicle automatic alignment method according to the present application includes:

[0039] S1 Identify the towing edge: Identify at least one towing edge of the trailer among the front edge and the rear edge in the traveling direction of the trailer;

[0040] S2 Calculate the actual position coordinates: According to the recognition result, calculate the actual position coordinates of one of the front edge and the rear edge;

[0041] S3 Determine the offset: Determine the offset of the towing relative to the conveyor belt according to the actual position coordinates and the preset docking position coordinates; and

[0042] S4 Move and align: Move the towing relative to the conveyor belt according to the offset to achieve alignment.

[0043] Among them, at least one towing edge of the trailer among the front edge and the rear edge in the traveling direction of the trailer can be identified by a laser device, a camera, etc. Here, the laser device mentioned in the present application includes the cases of using lidar and laser sensors alone or in combination. As long as a three-dimensional point cloud map can be established to construct a three-dimensional space model and provide the surrounding environment contour information, no specific limitation is made.

[0044] In the case of using a camera for recognition, through visual data processing

[0045] In the case of using a laser device for recognition, identifying the towing edge includes:

[0046] Scan a target scene including at least one trailer to form a point cloud of the target scene;

[0047] Perform target recognition and extraction on the point cloud of the target scene to identify and extract a point cloud set constituting at least one trailer from the point cloud;

[0048] Divide the point cloud set of at least one trailer into multiple point cloud subsets of the same shape along the driving direction of the trailer; and

[0049] Identify the point cloud subset constituting either the front edge or the rear edge by comparing the number of points in multiple point cloud subsets.

[0050] Specifically, in this embodiment, for example, a data processing unit can be used to perform the above point cloud data processing. After the original point cloud data of the scanned target scene is generated by a laser device, the original point cloud data received from the above laser device is parsed to obtain point cloud data that converts angle and distance information into three-dimensional coordinate XYZ values, and the external parameters of the point cloud data are adjusted based on the installation angle and installation height of the laser device to align the laser coordinate system and the world coordinate system.

[0051] Next, the point cloud is subjected to data filtering processing according to a pre-input height range to remove data outside the height range, forming the point cloud set constituting the at least one trailer.

[0052] Next, the point cloud set is sliced into multiple point cloud subsets arranged in a thin sheet shape (for example, the shape of a thin sheet of a cuboid) along the driving direction of the trailer.

[0053] Next, compare the number of points in the point cloud subsets. When the difference in the number of points between one point cloud subset and another point cloud subset in two adjacent point cloud subsets arranged among multiple point cloud subsets reaches a preset first threshold, it is determined that one of the two point cloud subsets is the point cloud subset constituting the front edge or the rear edge. More specifically, if the number of points in the previous point cloud subset arranged along the driving direction of the trailer is more than the number of points in the adjacent subsequent point cloud subset and exceeds the first threshold, it is judged as the rear edge of the trailer; if the number of points in the subsequent point cloud subset arranged along the driving direction of the trailer is more than the number of points in the adjacent previous point cloud subset and exceeds the first threshold, it is judged as the rear edge of the trailer.

[0054] In the calculation of the actual position coordinates, for example, a data processing unit can be used to calculate the actual position coordinates of either the front edge or the rear edge according to the recognition result of the trailer edge, specifically including:

[0055] Fitting a plane where one of the front edge or the rear edge is located according to the points in the point cloud subset that constitutes one of the front edge or the rear edge; and

[0056] Determining the actual position coordinates of one of the front edge or the rear edge according to the coordinate information of the fitted plane.

[0057] For example, the RANSAC algorithm can be used to fit the edge plane and calculate the actual position coordinates of the edge plane.

[0058] During the determination of the offset, for example, the data processing unit can be used to determine the offset of the trailer relative to the conveyor belt according to the actual position coordinates and the preset docking position coordinates;

[0059] During the moving alignment, for example, the vehicle management system can be used to move the trailer relative to the conveyor belt according to the offset to achieve alignment.

[0060] Next, embodiments will be further enumerated to illustrate the present invention in detail. It should also be understood that the following embodiments are only used to further illustrate the present invention and cannot be construed as limiting the protection scope of the present invention. Some non-essential improvements and adjustments made by those skilled in the art based on the above content of the present invention all fall within the protection scope of the present invention.

[0061] As Figure 2 shown, for example, on flat ground in scenarios such as docks, airport warehouses, and logistics warehouses, multiple conveyor belts with the same structure are arranged in a parallel manner and separated by a specified distance L. Figure 1 Only two conveyor belts are schematically marked in the figure, defined as conveyor belt 11 and conveyor belt 12 respectively. The transport vehicle has multiple trailers with the same model and connected end to end. Figure 1 Only two trailers are schematically marked in the figure, defined as trailer 21 and trailer 22 respectively. The front and rear trailers are separated by a specified interval D. The trailer itself is a rigid structure, and the connection between the trailers is a non-rigid connection such as a hinge connection. Trailer 21 and trailer 22 form part of the trailer. The trailer includes a towing vehicle head that towes these trailers and the above-mentioned trailers. Whether transporting goods from the trailer to the corresponding conveyor belt or from the conveyor belt to the corresponding trailer, it is necessary to extend the multiple trailers along the driving direction (i.e., the length direction) and laterally lean against one end of the corresponding conveyor belt, and at the same time, at least one of the multiple trailers stops at a predetermined position and is accurately aligned with the corresponding conveyor belt. Among them, generally, the conveying direction F1 of the belt of the conveyor belt is substantially perpendicular to the driving direction F2 of the trailer along the length. In addition, between two adjacent conveyor belts 11 and 12, at a height H from the ground and on the side where the trailer docks, a lidar 10 with a laser scanning range at least greater than the specified interval D is provided.

[0062] When the trailer is approaching, the lidar 10 scans the trailer and the surrounding environment to form point cloud data of the target scene. Specifically, the lidar 10 preferably has a scanning height range that is at least greater than the distance between the cargo panel of the trailer vehicle and the ground, a scanning width range that is at least greater than the specified interval D, and a scanning depth range that is at least greater than the width of the trailer. In addition, the lidar 10 here at least includes a transmitter, a receiver, a scanning mechanism, and a data processing module, and GPS can also be integrated when necessary to enhance the positioning ability and data accuracy.

[0063] Here, the installation height H of the lidar 10 can be, for example, approximately flush with the upper surface of the cargo panel of the trailer, and preferably has a downward inclination angle (i.e., the installation angle A) so as to scan the edge of the trailer and the ground at the same time. If the installation height H is much higher than the height of the upper surface of the cargo panel of the trailer (it can be imagined that the radar scans the trailer from above), there will be a situation where it is impossible to effectively distinguish the ground from the trailer, and there will also be a situation where the edge of a large-volume cargo extends outside the trailer, causing edge occlusion, both of which will result in the lidar 10 being unable to scan the true edge of the trailer and affecting edge positioning. On the other hand, when the installation height H is approximately flush with the height of the upper surface of the cargo panel of the trailer, the lidar 10 can effectively avoid the occlusion of the trailer edge by large-volume cargo as much as possible, and can also minimize noise points to distinguish the ground from the trailer well.

[0064] Here, the lidar 10 can detect only either the front edge or the rear edge of the target trailer, calculate the offset of the target trailer based on the three-dimensional coordinates of the front edge or the rear edge, or can also detect the front edge of the trailer located on the rear side in the driving direction (referred to as the rear-side trailer) and the rear edge of the trailer located on the front side in the driving direction (referred to as the front-side trailer) among two adjacent trailers in the driving direction, and as will be described in detail later, calculate the offset of the front-side trailer based on the three-dimensional coordinates of the rear edge of the front-side trailer, and calculate the offset of the rear-side trailer based on the three-dimensional coordinates of the front edge of the rear-side trailer. If both the front and rear edges are detected simultaneously, it can assist in verifying the accuracy of edge recognition. For example, when the distance between the separately recognized edges deviates significantly from the specified interval D, it indicates that at least one edge recognition is incorrect.

[0065] Next, the data processing unit analyzes the raw point cloud data received by the lidar 10 that cannot be directly used, and obtains point cloud data that converts the angle and distance information into three-dimensional coordinate XYZ values.

[0066] Next, if necessary, the extrinsic parameters of the point cloud data can be adjusted based on the installation angle A and installation height H of the lidar 10 to align the lidar coordinate system and the world coordinate system (also known as the global coordinate system). As described above, the point cloud coordinate system obtained by parsing the original point cloud belongs to the lidar coordinate system, and the coordinate system of the lidar needs to be transformed into the world coordinate system. Specifically, since the lidar is rigidly connected to the conveyor belt and the conveyor belt is fixed relative to the ground, during the alignment movement, the relative attitude and displacement between the lidar and the ground are fixed. Only by establishing the positional relationship between the two relative coordinate systems and through rotation or translation, can these two three-dimensional coordinate systems be unified into a single three-dimensional coordinate system.

[0067] Next, if necessary, the point cloud data can be filtered based on the installation height H of the lidar 10, etc. During the process of acquiring the point cloud data, due to the influence of factors such as the properties of the target object and / or the external environment, the point cloud data will inevitably be mixed with some noise points and / or outliers (noise points refer to the point cloud data that is useless for model processing, and outliers refer to the point cloud data that is far from the subjective measurement area. In this application, noise points include outliers), and they need to be directly removed or processed in a smoothed manner. For example, common filtering algorithms in the industry can be used, such as the average filtering algorithm, the first-order filtering algorithm, the Kalman filtering algorithm, etc.

[0068] Next, according to the pre-input height range, the point cloud is filtered to form a point cloud set of the trailer 21 and / or the trailer 22. Specifically, a large part of the point cloud data still belongs to the three-dimensional point data of the ground below the trailer or the trailer head or the three-dimensional point data of other objects (such as goods placed on the trailer) above the trailer in the height direction. This will affect the subsequent point cloud processing process of the target object. On the one hand, if these three-dimensional point data above or below the trailer are not removed, these redundant three-dimensional point data will interfere with the determination of the point cloud set of the trailer 21 and / or the trailer 22, and will reduce the accuracy and robustness of the trailer edge recognition. On the other hand, due to the large amount of point cloud data, this will increase the computational requirements of the model. Therefore, a height range representing the range of the trailer in the height direction is pre-input, and data filtering is performed based on this height range to remove all three-dimensional point data outside this height range. In this way, not only the amount of point cloud data to be processed is greatly reduced, but also the edge misjudgment caused by the edge of large-volume goods extending outside the trailer can be avoided. In addition, as an alternative to the height range, the upper surface (i.e., the cargo panel) or the lower surface of the trailer and the thickness of the trailer can also be pre-input, so that the set of point clouds constituting the trailer can also be screened to further streamline the data.

[0069] Next, for the point cloud set constituting the trailer, the point cloud set is sliced into multiple point cloud subsets in the shape of cuboids arranged along the driving direction of the trailer along the driving direction. In this embodiment, the point cloud subsets are cut into thin slices based on the shape of the target object, especially the edge. Specifically, the point cloud is cut into N sheet-shaped point cloud subsets along the driving direction F2, where N≥2. The length of the point cloud subset in the conveyor belt conveying direction F1 is, for example, 0.5 to 3 m, preferably the width of the trailer. The height in the direction perpendicular to the ground is, for example, 0.1 to 1 m, preferably the thickness of the trailer. The width in the driving direction F2 is, for example, 0.01 to 0.1 m, preferably 0.05 m (which can also be vividly understood as the thickness of the thin slice). Thus, by adopting the algorithm optimization mechanism of the reconfigurable multi-ring network, the efficiency of the algorithm operation is improved.

[0070] Next, by comparing the number of points in adjacent point cloud subsets, the point cloud subsets constituting either the front edge or the rear edge are identified. Specifically, the number of point clouds in each of the above cuboids is sequentially counted along the driving direction F2, and by locating the point cloud subsets with rapid changes in the point cloud set of the trailer, the point cloud subsets near the edge area of the trailer are found. Since there are entities inside the trailer edge, the number of point clouds received by reflection is large, and there are no entities outside the trailer edge (and noise points have been filtered and three-dimensional points constituting the ground or goods have been removed), so the number of point clouds received by reflection is small. Therefore, if the number of point clouds in the current point cloud subset is significantly more than that in the subsequent point cloud subset, it is judged as the rear edge of the trailer; if the number of point clouds in the current point cloud subset is significantly less than that in the subsequent point cloud subset, it is judged as the front edge of the trailer.

[0071] Next, based on the point cloud subsets where the trailer edges are identified, the actual position coordinates of the edges can be effectively calculated using the Random Sample Consensus (RANSAC) algorithm. The RANSAC algorithm is a model shape estimation algorithm used to estimate a predefined model and its corresponding parameters from a sampled data set, and it has good robustness. It establishes a plane equation through three randomly selected point clouds, substitutes the point cloud data into the plane equation in turn, and then determines whether the point is a point in the plane according to the set distance threshold. For example, points within the threshold range are inliers, and points outside the threshold are outliers, and through multiple iterations, the model equation is obtained. Specifically, the three-dimensional coordinates of the small plane of the front edge of the trailer are fitted by RANSAC plane, and these three-dimensional coordinates are also the position coordinates of the front edge of the trailer in the real scene. Since the trailer is a rigid structure, it is also equivalent to the position coordinates of the trailer in the real scene.

[0072] Next, based on the three-dimensional position coordinates of the trailer edges calculated above, the offset (offset distance) between the three-dimensional position coordinates and the pre-set (relative to the conveyor belt) docking position coordinates is further calculated.

[0073] Next, the vehicle scheduling management system (FMS) confirms the trailers to be aligned according to manual instructions, system-prestored instructions, etc. For example, via a communication system, etc., it obtains relevant information of the trailers to be aligned based on the lidar 10, such as the coordinate information of the edges and the offset, etc. If the lidar 10 only detects any one of the front and rear edges of the target trailer, the FMS can only obtain information for that trailer. If the lidar 10 simultaneously detects the front edge of the rear trailer and the rear edge of the front trailer among two adjacent trailers in the driving direction, the FMS can obtain information for any one of the two trailers. Thus, only one detection unit can freely select and control two trailers, making on-site scheduling more flexible and convenient. Here, the vehicle scheduling management system can, for example, schedule the driving route of the vehicle through indoor GPS positioning technology, etc., to ensure that the vehicle stops at a position closer to the conveyor belt.

[0074] Next, after receiving the required information (coordinate information of the edges, offset, etc.), the FMS performs corresponding processing, and uses vehicle networking technology (Vehicle to Everything; V2X), etc., to control the movement of the trailer so that the trailer moves to a preset docking position, realizing the precise alignment of the trailer and the conveyor belt.

[0075] In summary, in an embodiment of the present application, the edge of the trailer is laser scanned by a lidar, the front edge line segment of the trailer is fitted according to the filtered and segmented point cloud, and the 3D coordinates of the edge line segment are obtained. According to the 3D coordinates, the offset from the target position (including the offset distance, etc.) can be known, and then precise alignment is achieved. Therefore, according to the present application, the control accuracy of automatic vehicle alignment can be improved at a relatively low cost. More specifically, it only requires a lidar near the conveyor belt and the trailer to be able to scan the front edge or the rear edge of the trailer, without other structures, and the precise alignment of the trailer and the conveyor belt can be achieved, with the error remaining within centimeters. Moreover, during the entire alignment process, no manual intervention is required, there is no dependence on the experience and skills of the operator, etc., and there are no requirements for the types of trailers and conveyor belts. It can be adapted to most traditional trailers and conveyor belts on the market, and the position and attitude of the trailer can be accurately detected.

[0076] In addition, in an embodiment of the present application, the lidar is installed on the conveyor belt, but it is not limited thereto. As long as the above method can be implemented, the installation position is not restricted. In addition, relevant information of the conveyor belt can be pre-recorded in the lidar, so as to further obtain the positional relationship, relative angle, etc. of the trailer with respect to the conveyor belt. For example, after laser scanning and positioning the edge, by determining whether the edge is horizontal (such as comparing with a pre-stored horizontal model, etc.), it can be determined whether the ground (i.e., the trailer) is horizontal. If the ground is uneven, a feedback signal can be sent for scenario prompting or requesting to pause the alignment operation, etc. For example, after laser scanning and positioning the edge, by determining whether the edge is parallel to the conveying direction F1 of the conveyor belt (such as comparing with a pre-stored parallel model, etc.), it can be determined whether there is a deflection angle of the trailer in the traveling direction F2. If there is a deflection angle, a feedback signal can be sent for scenario prompting or requesting further adjustment of the alignment, etc.

[0077] In addition, the dimension in the width direction of the belt of each conveyor belt (i.e., perpendicular to the conveying direction F1) is the same as the dimension in the length direction of the cargo panel of each trailer (i.e., the traveling direction F2), and the same specified distance is separated between the front and rear trailers and between adjacent conveyor belts. Thus, after achieving precise alignment of one trailer with one conveyor belt, as long as it is ensured that there is no angular deviation of each trailer in the traveling direction F2, simultaneous alignment of multiple conveyor belts and multiple trailers can be achieved.

[0078] In addition, a computer-readable storage medium may include: any entity or device capable of carrying a computer program, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disc, a computer memory, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), and a software distribution medium, etc. The computer program includes computer program code. The computer program code may be in the form of source code, object code, an executable file, or some intermediate form, etc. A computer-readable storage medium may include: any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disc, a computer memory, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), and a software distribution medium, etc.

[0079] In addition, any process or method description depicted in the flowchart or described otherwise herein can be understood to represent a module, segment, or portion of code including one or more executable instructions for implementing a specific logical function or process. The scope of the preferred embodiments of the present application includes additional implementations, where functions can be executed in a substantially simultaneous manner or in a reverse order according to the functions involved, rather than in the order shown or discussed. This should be understood by those skilled in the art to which the embodiments of the present application pertain.

[0080] In addition, those of ordinary skill in the art can realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described according to their functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Skilled professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0081] Finally, it should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

Claims

1. A vehicle automatic alignment method, characterized in that: The vehicle automatic alignment method is a method for automatically aligning the trailer of the trailer and the conveyor belt, comprising: identifying one of a front edge and a rear edge of at least one hitch of the trailer in a direction of travel of the trailer; Calculate the actual position coordinates of one of the front edge and the rear edge according to the recognition result; Determining the offset of the trailer relative to the conveyor belt according to the actual position coordinates and the preset docking position coordinates; and The trailer is moved relative to the conveyor belt to align with the conveyor belt according to the offset.

2. The vehicle automatic alignment method according to claim 1, characterized in that: Identifying one of a front edge and a rear edge of at least one trailer of the trailer in a travel direction of the trailer comprises: Scanning a target scene including the at least one trailer by a laser radar to form a point cloud of the target scene; Performing target recognition and extraction on the point cloud of the target scene to recognize and extract a set of point clouds constituting the at least one trailer from the point cloud; Dividing the at least one trailer point cloud set into a plurality of point cloud subsets of the same shape along the driving direction of the trailer; and By comparing the numbers of points in adjacent point cloud subsets, a point cloud subset constituting one of the front edge or the rear edge is identified.

3. The vehicle automatic alignment method according to claim 2, characterized in that: When the laser radar is scanning, the received raw point cloud data is parsed to obtain point cloud data that converts angle and distance information into three-dimensional coordinate XYZ values, and the external parameters of the point cloud data are adjusted based on the installation angle and installation height of the laser radar to align the laser coordinate system and the world coordinate system.

4. The vehicle automatic alignment method according to claim 2, characterized in that: When target recognition and extraction are performed on the point cloud of the target scene, based on the pre-input height range, point cloud data outside the height range is removed to form the point cloud set.

5. The vehicle automatic alignment method according to claim 2, characterized in that: When the point cloud set is divided into a plurality of the point cloud subsets, the point cloud set is sliced ​​into a plurality of thin-sheet point cloud subsets arranged along the driving direction of the trailer.

6. The vehicle automatic alignment method according to claim 2 or 5, characterized in that: The point cloud subset is in the shape of a rectangular sheet with a length of 0.5 to 3 m, a height of 0.1 to 1 m, and a width of 0.01 to 0.1 m.

7. The vehicle automatic alignment method according to claim 2, characterized in that: When the difference between the number of points in one point cloud subset of two adjacently arranged point cloud subsets in a plurality of the point cloud subsets and the number of points in the other point cloud subset reaches a preset first threshold, it is determined that one of the two point cloud subsets is the point cloud subset constituting the front edge or the rear edge.

8. The vehicle automatic alignment method according to claim 7, characterized in that: If the number of points in the previous point cloud subset arranged along the driving direction of the trailer is greater than the number of points in the adjacent next point cloud subset and exceeds the first threshold, it is determined to be the rear edge of the trailer; If the number of points in the subsequent point cloud subset arranged along the driving direction of the trailer is greater than the number of points in the adjacent previous point cloud subset and exceeds the first threshold, it is determined to be the rear edge of the trailer.

9. The vehicle automatic alignment method according to claim 2, characterized in that: Calculating the actual position coordinates of one of the front edge and the rear edge according to the recognition result includes: Fitting a plane where one of the front edge or the rear edge is located according to points in a point cloud subset constituting one of the front edge or the rear edge; and The actual position coordinates of one of the front edge or the rear edge are determined according to the coordinate information of the fitted plane.

10. The vehicle automatic alignment method according to claim 9, characterized in that: The coordinate information of the plane of the front edge or the rear edge is fitted using the RANSAC algorithm.

11. The vehicle automatic alignment method according to claim 2, characterized in that: The laser radar is arranged between two adjacent conveyor belts in the traveling direction of the trailer, at a prescribed height from the ground and on a side close to the trailer; The laser scanning range of the laser radar is at least larger than the interval between adjacent trailers in the driving direction of the trailer.

12. The vehicle automatic alignment method according to claim 1, characterized in that: The trailer travels on a flat ground in one of a dock, an airport warehouse, and a logistics warehouse.

13. An electronic system, comprising a memory, a processor, and a computer program stored in the memory, wherein: The processor executes the computer program to implement the method of any one of claims 1 to 12.

14. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 12 is implemented.

15. A computer program product, comprising a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 12 is implemented.

Images