A multi-line laser radar-based equipment and method for parking commercial vehicles in an orderly manner

Abstract

A kind of device and method for parking commodity vehicles in order based on multi-line laser radar belong to the field of robot motion driving and control technology.The robot parking position is planned by cloud scheduling system, and the actual robot parking position is uploaded to cloud scheduling system to assist cloud scheduling system to plan the next robot parking position;The multi-line laser radar on the top of the robot is used to locate and orient the front parked commodity vehicle at a long distance, and the front parked commodity vehicle is aligned;The multi-line laser radar on the left and right sides in front of the robot is used to correct deviation at the middle section, and the appropriate distance with the left and right parked commodity vehicles is maintained;The multi-line laser radar on the left and right sides in front of the robot is used to detect the distance with the front parked commodity vehicle at a short distance, and the robot parking position is determined.The present application is suitable for robot motion driving, logistics and warehousing fields, and can improve the parking number of commodity vehicles in dense parking area, reduce labor intensity, and improve work efficiency and work accuracy.

Description

Technical Field

[0001] This invention relates to a method for neatly parking commercial vehicles in a dense parking area in the field of robot transfer technology, and more particularly to an equipment and method for neatly parking commercial vehicles based on multi-line lidar, belonging to the field of robot motion drive and control technology. Background Technology

[0002] With the advancement of smart port construction and the development of robotics, sensor technology, and 5G communication technology, mobile robots are being used more and more widely in ports. Automobile terminals handle huge volumes of traffic and need to store a large number of vehicles. Adopting a dense parking method for automobiles can maximize the utilization of the terminal's yard area and increase the number of vehicles that can be parked. Currently, dense parking of automobiles mainly relies on manual labor, resulting in a significant waste of human resources and low work efficiency. The high cost and diverse sizes of automobiles require dense parking with high precision and greater flexibility; currently, there are no known cases of using mobile robots for dense parking of automobiles. Summary of the Invention

[0003] To address the problems of low parking efficiency and high labor costs in densely packed parking areas of automobile terminals, the main objective of this invention is to provide equipment and methods for neatly parking automobiles based on multi-line LiDAR. The multi-line LiDAR, mounted on a automobile transfer robot, collects environmental point cloud information; and by processing this point cloud information, the robot's movement path is corrected to determine the final parking position. This increases the number of automobiles that can be parked in densely packed areas, reduces labor intensity, and improves work efficiency and accuracy.

[0004] The objective of this invention is achieved through the following technical solution:

[0005] An equipment for neatly parking commercial vehicles based on multi-line LiDAR comprises a cloud-based scheduling system, a commercial vehicle transfer robot, a clamping system, an environmental perception system, and a robot control system. The cloud-based scheduling system contains map information of the automobile terminal, used to issue vehicle retrieval and placement commands to the commercial vehicle transfer robot and plan the global path. The commercial vehicle transfer robot is a mobile robot used for transferring commercial vehicles. The clamping system is mounted on the commercial vehicle transfer robot and used to clamp the commercial vehicles to be transferred. The environmental perception system includes a 3D servo turntable mounted on the top front of the transfer robot, multi-line LiDAR mounted on the 3D servo turntable, and multi-line LiDARs mounted on both sides of the front of the commercial vehicle transfer robot. The 3D servo turntable can rotate freely at multiple angles, increasing the detection range of the multi-line LiDARs. The multi-line LiDARs mounted on the 3D servo turntable are used for long-range scanning of commercial vehicles. The multi-line LiDARs mounted on both sides of the front of the commercial vehicle transfer robot can scan the commercial vehicles loaded in front, to the left and right, and inside the robot at close range. The robot control system is used to control the operation of the commercial vehicle transfer robot, communicate with the cloud-based scheduling system, and process the information collected by the environmental perception system.

[0006] A method for neatly parking commercial vehicles based on multi-line lidar, implemented using equipment for neatly parking commercial vehicles based on multi-line lidar, includes the following steps:

[0007] Step 1: The dock operating system sends vehicle placement task information for the confined parking area to the cloud scheduling system. This task information includes the vehicle dimensions, retrieval area location, and the placement area location within the confined parking area. The vehicle transfer robot uploads its vehicle-side information to the cloud scheduling system, including its current location. If the confined parking area is being planned for the first time, the cloud scheduling system plans the vehicle transfer robot's current placement and retrieval locations based on the confined parking area placement task information. If the confined parking area is not being planned for the first time, the cloud scheduling system plans the current placement and retrieval locations based on the confined parking area placement task information and the vehicle transfer robot's previous placement location. The cloud scheduling system uses the vehicle-side information, placement location, and retrieval location to plan the global retrieval and global placement paths for the vehicle transfer robot, and then sends these paths to the vehicle transfer robot.

[0008] Step 2: The vehicle transfer robot receives the global vehicle retrieval path and global vehicle release path from the cloud scheduling system, arrives at the vehicle retrieval location along the global vehicle retrieval path, and uses the clamping mechanism to load the vehicle to be transferred.

[0009] Step 3: The vehicle transfer robot moves along the global vehicle placement path. When it is at the first set distance from the end of the global vehicle placement path, the robot control system plans a new execution path based on the environmental point cloud information collected by the multi-line lidar installed on the top front of the vehicle transfer robot.

[0010] 3.1 When the vehicle transfer robot reaches a position at a predetermined distance from the end point of the global vehicle placement path, the robot stops moving. The robot control system controls the 3D servo turntable to perform a downward head-swing motion at a predetermined angle. During the head-swing of the 3D servo turntable, the multi-line LiDAR installed on the 3D servo turntable continuously collects environmental point cloud information from top to bottom. After the 3D servo turntable completes the head-swing, the robot control system accumulates the environmental point cloud information collected by the multi-line LiDAR and performs range filtering within a predetermined range, retaining only the environmental point cloud information in front of the vehicle transfer robot.

[0011] 3.2. The point cloud information of the environment in front of the vehicle transfer robot collected from top to bottom in step 3.1 contains vehicle information. Then, by processing the point cloud information of the environment, a new execution path for the vehicle transfer robot is planned.

[0012] 3.2.1 Within the set angle range, traverse all angles in 1-degree increments. Obtain the bounding box, area, and center coordinates of the vehicle point cloud for each angle direction.

[0013] 3.2.2 Select the rectangle with the smallest area, and use the corresponding angle as the heading angle of the vehicle in front of the vehicle transfer robot. Use the center position of the rectangle with the smallest area as the position coordinate of the vehicle in front of the vehicle transfer robot.

[0014] 3.2.3 The robot control system plans a straight path based on the position coordinates of the vehicles in front of the vehicle transfer robot as the final release path, and replaces the global release path issued by the cloud scheduling system as the new execution path.

[0015] 3.3 If the environmental point cloud information in front of the transfer robot obtained in step 3.1 does not contain information about the goods vehicle, then the global vehicle release path issued by the cloud scheduling will be used as the new execution path.

[0016] Step 4: The vehicle transfer robot moves along the new execution path obtained in Step 3. Two multi-line lidars installed on both sides of the front end of the vehicle transfer robot obtain the distance between the vehicle inside the robot and the vehicles parked on the left and right, thereby correcting the movement of the vehicle transfer robot.

[0017] 4.1 Two multi-line lidars installed on both sides of the front end of the vehicle transfer robot scan the surrounding environment and acquire environmental point cloud data. The environmental point cloud data acquired by the two multi-line lidars are merged and range filtering within a set range is used to obtain point cloud data of the inside, front, left and right sides of the vehicle transfer robot, thereby obtaining environmental information of the vehicle transfer robot in various directions.

[0018] 4.2 Merge the point clouds inside, to the left and to the right of the vehicle transfer robot obtained in step 4.1 to obtain the environmental point cloud information of the vehicle transfer robot in the lateral direction;

[0019] 4.3. If the environmental point cloud information of the vehicle transfer robot in the lateral direction obtained in step 4.2 contains vehicle information, then the environmental point cloud information is processed to obtain the distance between the vehicle already loaded inside the cavity of the vehicle transfer robot and the vehicles already parked on the left and right. The correction amount is obtained from the distance and added to the wheel angle of the vehicle transfer robot so that the vehicle inside the vehicle transfer robot maintains a set distance from the vehicles already parked on the left or right.

[0020] 4.3.1 For the point cloud after compression and merging in the lateral direction of the vehicle transfer robot, only the coordinate values ​​of the merged point cloud in the lateral direction of the vehicle transfer robot are retained.

[0021] 4.3.2 Sort the coordinate values ​​from largest to smallest, and calculate whether the difference between each pair of adjacent coordinate values ​​is greater than the set distance threshold. If it is greater than the threshold, it means that the point cloud of the current coordinate value and the point cloud of the previous coordinate value do not belong to the same category in the lateral direction of the commodity vehicle transfer robot; otherwise, they belong to the same category.

[0022] 4.3.3 Remove all classes whose point cloud count is less than the set minimum point cloud count.

[0023] 4.3.4 One of the point cloud classes represents the vehicle carried inside the vehicle transfer robot. If other point cloud classes exist, it means that there are parked vehicles to the left and right of the vehicle transfer robot; if no other point cloud classes exist, it means that there are no parked vehicles to the left and right of the vehicle transfer robot.

[0024] 4.4 If the environmental point cloud information of the vehicle transfer robot in the lateral direction obtained in step 4.2 does not contain vehicle information, the vehicle transfer robot will continue to move along the new execution path obtained in step 3.

[0025] Step 5: When the vehicle transfer robot reaches the second set distance from the end of the path, the multi-line laser radar installed at both ends in front of the vehicle transfer robot obtains the distance between the vehicle inside the vehicle transfer robot and the vehicles already parked on the left and right, thereby determining the parking position of the vehicle transfer robot.

[0026] 5.1 Merge the point clouds inside and in front of the vehicle transfer robot obtained in step 4.1 to obtain the environmental point cloud information of the vehicle transfer robot in the longitudinal direction.

[0027] 5.2. If the environmental point cloud information in the longitudinal direction of the vehicle transfer robot obtained in step 5.1 contains vehicle information, then by processing the environmental point cloud information, the distance between the vehicle carried inside the vehicle transfer robot and the vehicle parked in front can be obtained, thereby determining the parking position of the vehicle transfer robot.

[0028] 5.2.1 For the point cloud after compression and merging in the longitudinal direction of the vehicle transfer robot, only the coordinate values ​​of the merged point cloud in the longitudinal direction of the vehicle transfer robot are retained.

[0029] 5.2.2 Sort the coordinate values ​​from largest to smallest, and calculate whether the difference between each pair of adjacent coordinate values ​​is greater than the set distance threshold. If it is greater than the threshold, it means that the point cloud of the current coordinate value and the point cloud of the previous coordinate value do not belong to the same category in the lateral direction of the vehicle transfer robot; otherwise, they belong to the same category.

[0030] 5.2.3 Remove classes with a point cloud count less than the set minimum point cloud count. One of the point cloud classes represents the vehicles carried inside the vehicle transfer robot. If other point cloud classes exist, they represent vehicles in front of the vehicle transfer robot; if no other point cloud classes exist, it means there are no parked vehicles in front of the vehicle transfer robot.

[0031] 5.2.4 Calculate the current distance between the vehicle inside the vehicle transfer robot and the vehicle in front, and subtract it from the distance between the vehicles in front and behind the required close-stop zone to obtain the distance the vehicle transfer robot needs to continue moving. After controlling the vehicle transfer robot to move the required distance, control the vehicle transfer robot to stop moving.

[0032] 5.3 If the environmental point cloud information in the longitudinal direction of the vehicle transfer robot obtained in step 5.1 does not contain vehicle information, then when the vehicle transfer robot reaches the end of the global vehicle placement path, the vehicle transfer robot control system will control the vehicle transfer robot to stop moving.

[0033] Step 6: The vehicle transfer robot control system controls the clamping mechanism to complete the vehicle placement operation. It uses a positioning system to obtain the vehicle's location coordinates and uploads these coordinates to the cloud-based scheduling system, completing one placement operation. This increases the number of vehicles that can be parked in densely packed parking areas, reduces labor intensity, and improves work efficiency and accuracy.

[0034] Beneficial effects:

[0035] 1. The present invention discloses an equipment and method for neatly parking commercial vehicles based on multi-line lidar. It adopts a cloud-guided vehicle-end and vehicle-end feedback to cloud. The cloud plans the parking position, and after the vehicle-end reaches the parking position, it feeds back to the cloud to assist the cloud in planning the parking position of the next commercial vehicle. This enables the commercial vehicle transfer robot to coordinate planning through the cloud and the vehicle end, ensuring that the commercial vehicle transfer robot parks the vehicles neatly.

[0036] 2. The present invention discloses an equipment and method for neatly parking commercial vehicles based on multi-line lidar. After the commercial vehicle transfer robot arrives at the dense parking area, it autonomously determines the parking position without relying on cloud data. It locates and orients the commercial vehicles already parked in front from a distance, corrects deviation based on the commercial vehicles already parked on the left and right sides in the middle section, and controls parking based on the distance to the commercial vehicles already parked in front at close range. This method ensures that the commercial vehicle transfer robot maintains a specified distance between the commercial vehicles parked by the robot and the surrounding commercial vehicles, while also aligning with the commercial vehicles in front, further guaranteeing the neatness of the commercial vehicles parked in the dense parking area.

[0037] 3. The present invention discloses an equipment and method for neatly parking commercial vehicles based on multi-line lidar. The environmental point cloud information scanned by two multi-line lidars installed on both sides in front of the commercial vehicle transfer robot is merged and range filtered to obtain environmental information inside, in front of, and on the left and right sides of the commercial vehicle transfer robot. This enables the division of the range of the environment around the commercial vehicle transfer robot, thereby classifying the positions of the commercial vehicles parked in front of and on the left and right sides of the commercial vehicle transfer robot and improving the accuracy of commercial vehicle parking.

[0038] 4. The present invention discloses an equipment and method for neatly parking commercial vehicles based on multi-line lidar. The acquired point cloud data is compressed in a specified direction to convert the point cloud data into one dimension and obtain coordinate values. By judging whether the distance difference between adjacent coordinate values ​​is greater than a set value, and since the commercial vehicle is relatively large, the class with less than a set value of point cloud is further eliminated. This enables point cloud analysis in a specific direction, reduces the computational burden, and improves parking efficiency.

[0039] 5. This invention discloses an equipment and method for neatly parking commercial vehicles based on multi-line lidar. In a densely packed parking area, the distance between two commercial vehicles is required to be very small. When the distance between a commercial vehicle inside the transfer robot and the commercial vehicle in front is too close, the two multi-line lidars installed on both sides in front of the transfer robot are not positioned between the commercial vehicle inside the transfer robot and the commercial vehicle in front. The multi-line lidars cannot clearly scan the front of the commercial vehicle in front of the transfer robot. By scanning the distance between the commercial vehicle inside the transfer robot and the commercial vehicle in front from a distance, the difference between the current distance and the target distance of the transfer robot can be controlled, thereby achieving a precise distance between the commercial vehicle inside the transfer robot and the commercial vehicle in front. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of a densely packed parking area at a car terminal, illustrating an equipment and method for neatly parking commercial vehicles based on multi-line lidar disclosed in this embodiment.

[0041] Figure 2 This is a schematic diagram of the structure of a vehicle-mounted parking system based on a multi-line lidar disclosed in this embodiment;

[0042] Figure 3 This is a flowchart of an equipment and method for neatly parking commercial vehicles based on multi-line lidar disclosed in this invention;

[0043] Figure 4 This is a flowchart of the clustering algorithm in a device and method for neatly parking commercial vehicles based on multi-line lidar disclosed in this embodiment;

[0044] Wherein: 1—A 3D servo turntable installed at the top front of the vehicle transfer robot; 2—A 32-line LiDAR installed on the 3D servo turntable; 3—A 32-line LiDAR installed on both sides of the front end of the vehicle transfer robot. Detailed Implementation

[0045] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. The technical problems solved by the present invention and its beneficial effects are also described. It should be noted that the described embodiments are only intended to facilitate understanding of the present invention and do not constitute any limitation thereof.

[0046] This embodiment discloses an equipment and method for neatly parking commercial vehicles based on multi-line lidar. In a densely packed parking area at a car terminal, it enables a vehicle transfer robot to neatly park vehicles, improving the space utilization rate of the densely packed parking area. (The text then abruptly shifts to a seemingly unrelated topic: "Car terminal densely packed parking area...") Figure 1 As shown, the required spacing between the front, back, left, and right sides of the vehicle is 30cm.

[0047] This embodiment discloses a vehicle-mounting device based on multi-line LiDAR, installed on a vehicle-transfer robot. The vehicle-transfer robot is 6m long and 3m wide, containing an internal cavity 6m long and 2.1m wide to hold the vehicles. The cavity has a clamping mechanism that grips and lowers the vehicle after docking. The vehicle-transfer robot uses a four-wheel independent drive steering mechanism. The device includes a three-dimensional servo turntable mounted on the top front of the vehicle-transfer robot, a 32-line LiDAR mounted on the turntable, and two 32-line LiDARs mounted on either side of the front of the vehicle-transfer robot. The installation positions are as follows: Figure 2 As shown.

[0048] This embodiment discloses a method for neatly parking commercial vehicles based on multi-line lidar, implemented using equipment for neatly parking commercial vehicles based on multi-line lidar, such as... Figure 3 As shown, the specific implementation steps are as follows:

[0049] Step 1: The cloud-based scheduling system plans the vehicle placement location, global vehicle retrieval path, and global vehicle placement path for the vehicle transfer robot, and then sends these plans to the vehicle transfer robot.

[0050] 1.1 The dock operating system sends a vehicle release task instruction to the cloud dispatch system. The vehicle release task instruction includes the location information of the vehicle retrieval area, the location information of the vehicle release area in the dense parking area, and the specifications and dimensions of the vehicle.

[0051] 1.2 If this is the first time a vehicle is released from the dense parking area, the cloud-based dispatch system will directly plan the release location of the vehicle transfer robot based on the release task information. If this is not the first time a vehicle is released from the dense parking area, the cloud-based dispatch system will plan the release location of the vehicle transfer robot based on the release task information and the previous release location of the vehicle transfer robot.

[0052] 1.3 The vehicle transfer robot uploads its location information to the cloud-based scheduling system. Based on the vehicle transfer robot's location information and the vehicle terminal map information, the cloud-based scheduling system plans the global vehicle retrieval path and global vehicle release path for the vehicle transfer robot, and sends the global vehicle retrieval path and global vehicle release path to the vehicle transfer robot using MQTT communication.

[0053] Step 2: The vehicle transfer robot receives the global vehicle retrieval path and global vehicle release path from the cloud scheduling system, arrives at the vehicle retrieval location along the global vehicle retrieval path, and uses the clamping mechanism to load the vehicle to be transferred.

[0054] Step 3: The vehicle transfer robot uses the GPS positioning system to obtain the current location of the robot at all times. When it reaches a position 7m away from the end point of the vehicle placement position, the vehicle transfer robot control system uses the environmental point cloud information collected by the multi-line laser radar installed on the top front of the vehicle transfer robot to plan a new execution path.

[0055] 3.1 When the vehicle transfer robot reaches a position 7m away from the end point of the vehicle placement location, the vehicle transfer robot stops moving, and the robot control system controls the three-dimensional servo turntable installed on the top front of the vehicle transfer robot to make a downward 60-degree head-down movement.

[0056] 3.2 During the downward movement of the 3D servo turntable, the 32-line lidar installed on the 3D servo turntable continuously scans the surrounding environment from top to bottom to obtain point cloud information of the surrounding environment, and accumulates multiple frames of point cloud information to obtain multi-angle point cloud data.

[0057] 3.3. Perform range filtering on the point cloud information of the surrounding environment scanned by the 32-line lidar installed on the 3D servo turntable, and retain only the point cloud information of the area 10m in front of the commodity vehicle transfer robot and 3m wide.

[0058] 3.4 If the point cloud information after range filtering in step 3.3 contains a vehicle, the vehicle transfer robot control system processes the point cloud after range filtering in step 3.3, plans a new execution path for the vehicle transfer robot, and controls the robot to move along the new execution path. The point cloud processing is as follows:

[0059] 3.4.1 Within the set angle range, traverse all angles in 1-degree increments. Obtain the bounding box, area, and center coordinates of the vehicle point cloud for each angle direction.

[0060] 3.4.2 Select the rectangle with the smallest area, and use the corresponding angle as the heading angle of the vehicle in front of the vehicle transfer robot. Use the center position of the rectangle with the smallest area as the position coordinate of the vehicle in front of the vehicle transfer robot.

[0061] 3.4.3 The robot control system plans a straight path based on the position coordinates of the vehicles in front of the vehicle transfer robot as the final release path, and replaces the global release path issued by the cloud scheduling system as the new execution path.

[0062] 3.5 If the point cloud information after range filtering in step 3.3 does not contain the vehicle, the vehicle transfer robot retains the global vehicle retrieval path issued by the cloud scheduling and moves according to the global vehicle retrieval path issued by the cloud scheduling.

[0063] Step 4: During the movement of the vehicle transfer robot, the 32-line LiDAR on both sides of the front end of the vehicle transfer robot constantly scans the point cloud information of the environment around the vehicle transfer robot, divides the point cloud into blocks, and outputs the point cloud data of the front, left, right and inside of the vehicle transfer robot respectively. The range of the front point cloud data is 20m long and 2.1m wide, the range of the left and right point cloud data is 20m long and 6m wide, and the range of the internal point cloud data is 6m long and 2.1m wide.

[0064] Step 5: Merge the point clouds inside, to the left and to the right of the vehicle transfer robot to obtain the environmental point cloud information of the vehicle transfer robot in the lateral direction;

[0065] 5.1 When the vehicle transfer robot scans a vehicle in the lateral direction, the vehicle transfer robot control system processes the point cloud data in the lateral direction to obtain the distance between the vehicles parked on the left and right sides of the vehicle transfer robot and the vehicles already loaded inside the vehicle transfer robot. The system then uses these distances to obtain a correction amount and adds it to the wheel angle of the vehicle transfer robot. This ensures that the vehicles inside the vehicle transfer robot maintain a set distance from the vehicles already parked on the left or right. The point cloud processing is as follows:

[0066] 5.1.1 Define the lateral direction of the vehicle transfer robot as the y-direction. Compress the point cloud data of the lateral direction of the vehicle transfer robot obtained in step 7 in the y-direction to obtain the coordinate values ​​of the point cloud data in the y-direction.

[0067] 5.1.2 Sort the point cloud data in descending order of the y-axis coordinates. Iterate through all coordinate values ​​and check if the difference between two adjacent coordinate values ​​is greater than the set distance threshold of 0.05m. If it is greater than the threshold, it means that the point cloud of the current coordinate value and the point cloud of the adjacent coordinate value do not belong to the same class. Further check if the number of points in the current class is greater than the set minimum number of points of 40. If the condition is met, keep it; otherwise, do not keep it.

[0068] 5.1..3 Select the two largest classes and calculate the distance between the two largest classes. This distance is defined as the distance between the left and right side of the goods vehicle and the internal goods vehicle of the goods vehicle transfer robot.

[0069] 5.2 If the vehicle transfer robot does not scan a vehicle on its left or right, it will continue moving according to the final vehicle placement path planned by the vehicle transfer robot.

[0070] Step 6: Merge the point clouds inside and in front of the vehicle transfer robot to obtain the environmental point cloud information in the longitudinal direction of the vehicle transfer robot;

[0071] 6.1 If the filtered point cloud information from step 3 contains a vehicle, the environmental point cloud information in the longitudinal direction of the vehicle transfer robot obtained in step 10 will also contain the vehicle. The environmental point cloud in the longitudinal direction of the vehicle transfer robot is processed to obtain the distance between the parked vehicle in front of the robot and the vehicle inside the robot. When the distance drops to 1m, the robot control system controls the vehicle transfer robot to move 0.7m and then stop. The environmental point cloud processing in the longitudinal direction of the vehicle transfer robot is as follows:

[0072] 6.1.1 Define the longitudinal direction of the vehicle transfer robot as the x-direction, merge the point cloud data in front of the vehicle transfer robot and the point cloud data inside the robot, compress the merged point cloud data in the x-direction, and obtain the coordinate value of the point cloud data in the x-direction.

[0073] 6.1.2 Sort the point cloud data by coordinate values ​​in the x-direction from largest to smallest. Traverse all coordinate values ​​and check if the difference between two adjacent coordinate values ​​is greater than the set distance threshold of 0.05m. If it is greater than the threshold, it means that the point cloud of the current coordinate value and the point cloud of the adjacent coordinate value do not belong to the same class. Further check if the number of points in the current class is greater than the set minimum number of points of 40. If the condition is met, the class is retained; if the condition is not met, the class is not retained.

[0074] 6.1.3 Select the two largest classes and calculate the distance between the two largest classes. This distance is defined as the distance between the goods vehicle inside the goods vehicle transfer robot and the goods vehicle in front.

[0075] 6.1.4 When the distance between the car inside the car transfer robot and the car in front of the car transfer robot reaches 1m, the robot control system controls the car transfer robot to move forward 0.7m and then stop.

[0076] 6.2 If the point cloud information after range filtering in step 3 does not contain the goods vehicle, the environmental point cloud information in the longitudinal direction of the goods vehicle transfer robot obtained in step 10 will not contain the goods vehicle. When the goods vehicle transfer robot moves to the end of the path, the robot is controlled to stop moving.

[0077] Step 7: The vehicle transfer robot control system controls the clamping mechanism to complete the vehicle placement operation. After the vehicle is placed, the GPS positioning system is used to obtain the location coordinates of the parked vehicle, and the location coordinates of the parked vehicle are uploaded to the cloud scheduling system via MQTT communication to update the location of the next parked vehicle. The vehicle transfer robot uses autonomous navigation technology to move the next vehicle.

[0078] By using cloud-based planning to determine the placement location and route of the vehicle transfer robot, and sending the plan to the vehicle, the robot can move along the route. At the final placement point, without relying on vehicle-cloud interaction, three multi-line LiDARs installed in different locations can detect the surrounding parked vehicles. The robot can be positioned and oriented based on parked vehicles at a long distance, correct its course based on parked vehicles on the left and right in the middle section, and control the robot to stop at close range based on parked vehicles. This allows the robot to neatly park vehicles in the dense parking area of ​​the car terminal.

[0079] The above detailed description further illustrates the purpose, technical solution, and beneficial effects of the invention. It should be understood that the above description is merely a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for neatly parking commercial vehicles based on multi-line lidar, characterized in that: Includes the following steps, Step 1: The dock operating system sends vehicle placement task information for the dense parking area to the cloud dispatch system. The task information includes the vehicle size, retrieval area location, and dense parking area placement area location. The vehicle transfer robot uploads its vehicle-side information to the cloud dispatch system, which includes the robot's current location. If the dense parking area is being planned for the first time, the cloud dispatch system plans the vehicle placement and retrieval locations for the vehicle transfer robot based on the placement task information. If the dense parking area is not being planned for the first time, the cloud dispatch system plans the placement and retrieval locations for the vehicle transfer robot based on the placement task information and the robot's previous placement location. The cloud-based scheduling system uses vehicle information, vehicle placement location, and vehicle retrieval location to plan the global retrieval path and global placement path of the commercial vehicle transfer robot, and then sends the global retrieval path and global placement path to the commercial vehicle transfer robot. Step 2: The vehicle transfer robot receives the global vehicle retrieval path and global vehicle release path from the cloud scheduling system, reaches the vehicle retrieval location along the global vehicle retrieval path, and uses the clamping mechanism to load the vehicle to be transferred. Step 3: The vehicle transfer robot moves along the global vehicle placement path. When it is at the first set distance from the end of the global vehicle placement path, the robot control system plans a new execution path based on the environmental point cloud information collected by the multi-line lidar installed on the top front of the vehicle transfer robot. Step 4: The vehicle transfer robot moves along the new execution path obtained in Step 3. Two multi-line lidars installed on both sides of the front end of the vehicle transfer robot obtain the distance between the vehicle inside the robot and the vehicles already parked on the left and right, thereby correcting the movement of the vehicle transfer robot. Step 5: When the vehicle transfer robot reaches the second set distance from the end of the path, the multi-line laser radar installed at both ends in front of the vehicle transfer robot obtains the distance between the vehicle inside the vehicle transfer robot and the vehicles parked on the left and right, thereby determining the parking position of the vehicle transfer robot. Step 6: The vehicle transfer robot control system controls the clamping mechanism to complete the vehicle release operation, uses the positioning system to obtain the position coordinates of the vehicle release, and uploads the position coordinates of the vehicle release to the cloud scheduling system to complete one vehicle release operation. The equipment implementing the method comprises a cloud-based scheduling system, a vehicle transfer robot, a clamping system, an environmental perception system, and a robot control system. The cloud-based scheduling system contains map information of the automobile terminal and is used to issue vehicle pick-up and release commands to the vehicle transfer robot and plan the global path. The vehicle transfer robot is a mobile robot used for transferring vehicles. The clamping system is mounted on the vehicle transfer robot and is used to clamp vehicles to be transferred. The environmental perception system includes a 3D servo turntable mounted on the top front of the vehicle transfer robot, multi-line LiDAR mounted on the 3D servo turntable, and multi-line LiDAR mounted on both sides of the front of the vehicle transfer robot. The 3D servo turntable can rotate freely at multiple angles, increasing the detection range of the multi-line LiDAR. The multi-line LiDAR mounted on the 3D servo turntable is used for long-distance scanning of vehicles. The multi-line LiDAR mounted on both sides of the front of the vehicle transfer robot can scan vehicles loaded in front, to the left and right, and inside the robot at close range. The robot control system is used to control the operation of the vehicle transfer robot, communicate with the cloud-based scheduling system, and process the information collected by the environmental perception system.

2. The method for neatly parking commercial vehicles based on multi-line lidar as described in claim 1, characterized in that: The implementation method for step 3 is as follows: 3.1 When the vehicle transfer robot reaches a position that is a first set distance from the end point of the global vehicle placement path, the robot stops moving; the robot control system controls the three-dimensional servo turntable to swing downwards at a set angle; during the swinging of the three-dimensional servo turntable, the multi-line LiDAR installed on the three-dimensional servo turntable continuously collects environmental point cloud information from top to bottom; after the three-dimensional servo turntable completes the swing, the robot control system accumulates the environmental point cloud information collected by the multi-line LiDAR and performs range filtering within a set range, retaining only the environmental point cloud information in front of the vehicle transfer robot; 3.

2. The point cloud information of the environment in front of the vehicle transfer robot collected from top to bottom in step 3.1 contains vehicle information. Then, by processing the point cloud information of the environment, a new execution path for the vehicle transfer robot is planned. 3.2.1 Within the set angle range, traverse all angles with a step size of 1 degree; obtain the bounding box, area of ​​the bounding box, and coordinates of the center position of the bounding box of the point cloud of the vehicle in each angle direction; 3.2.2 Select the rectangle with the smallest area, and use the corresponding angle as the heading angle of the vehicle in front of the vehicle transfer robot. Use the center position of the rectangle with the smallest area as the position coordinate of the vehicle in front of the vehicle transfer robot. 3.2.3 The robot control system plans a straight path based on the position coordinates of the vehicles in front of the vehicle transfer robot as the final release path, and replaces the global release path issued by the cloud scheduling system as the new execution path. 3.3 If the environmental point cloud information in front of the transfer robot obtained in step 3.1 does not contain information about the goods vehicle, then the global vehicle release path issued by the cloud scheduling will be used as the new execution path.

3. The method for neatly parking commercial vehicles based on multi-line lidar as described in claim 2, characterized in that: The implementation method for step 4 is as follows: 4.1 Two multi-line lidars installed on both sides of the front end of the vehicle transfer robot scan the surrounding environment and acquire environmental point cloud data. The environmental point cloud data acquired by the two multi-line lidars are merged and range filtering within a set range is used to obtain point cloud data of the inside, front, left and right sides of the vehicle transfer robot, thereby obtaining environmental information of the vehicle transfer robot in various directions. 4.2 Merge the point clouds inside, left and right of the vehicle transfer robot obtained in step 4.1 to obtain the environmental point cloud information of the vehicle transfer robot in the lateral direction; 4.

3. If the environmental point cloud information of the vehicle transfer robot in the lateral direction obtained in step 4.2 contains vehicle information, then the environmental point cloud information is processed to obtain the distance between the vehicle already loaded inside the cavity of the vehicle transfer robot and the vehicles already parked on the left and right. The correction amount is obtained from the distance and added to the wheel angle of the vehicle transfer robot so that the vehicle inside the vehicle transfer robot maintains a set distance from the vehicles already parked on the left or right. 4.3.1 For the point cloud after compression and merging in the lateral direction of the vehicle transfer robot, only the coordinate values ​​of the merged point cloud in the lateral direction of the vehicle transfer robot are retained. 4.3.2 Sort the coordinate values ​​from largest to smallest, and calculate whether the difference between each pair of adjacent coordinate values ​​is greater than the set distance threshold. If it is greater than the threshold, it means that the point cloud of the current coordinate value and the point cloud of the previous coordinate value do not belong to the same category in the lateral direction of the vehicle transfer robot; otherwise, they belong to the same category. 4.3.3 Remove all classes whose point cloud count is less than the set minimum point cloud count; 4.3.4 One of the point cloud classes represents the vehicle carried inside the vehicle transfer robot; if other point cloud classes exist, it means that there are parked vehicles to the left and right of the vehicle transfer robot. If there are no other point cloud types, there are no parked vehicles to the left or right of the vehicle transfer robot; 4.4 If the environmental point cloud information of the vehicle transfer robot in the lateral direction obtained in step 4.2 does not contain vehicle information, the vehicle transfer robot will continue to move along the new execution path obtained in step 3.

4. The method for neatly parking commercial vehicles based on multi-line lidar as described in claim 3, characterized in that: the implementation method of step 5 is as follows: 5.1 Merge the point clouds inside and in front of the vehicle transfer robot obtained in step 4.1 to obtain the environmental point cloud information of the vehicle transfer robot in the longitudinal direction; 5.

2. If the environmental point cloud information in the longitudinal direction of the vehicle transfer robot obtained in step 5.1 contains vehicle information, then by processing the environmental point cloud information, the distance between the vehicle carried inside the vehicle transfer robot and the vehicle parked in front can be obtained, thereby determining the parking position of the vehicle transfer robot. 5.2.1 For the point cloud after compression and merging in the longitudinal direction of the vehicle transfer robot, only the coordinate values ​​of the merged point cloud in the longitudinal direction of the vehicle transfer robot are retained. 5.2.2 Sort the coordinate values ​​from largest to smallest, and calculate whether the difference between each pair of adjacent coordinate values ​​is greater than the set distance threshold. If it is greater than the threshold, it means that the point cloud of the current coordinate value and the point cloud of the previous coordinate value do not belong to the same category in the lateral direction of the vehicle transfer robot; otherwise, they belong to the same category. 5.2.3 Remove classes with a point cloud count less than the set minimum point cloud count; one of the point cloud classes represents the vehicles carried inside the vehicle transfer robot; if other point cloud classes exist, they represent the vehicles in front of the vehicle transfer robot. If no other point cloud type exists, it means that there are no parked vehicles in front of the vehicle transfer robot. 5.2.4 Calculate the current distance between the vehicle inside the vehicle transfer robot and the vehicle in front, and subtract it from the distance between the vehicles in front and behind the required close-stop zone to obtain the distance the vehicle transfer robot needs to continue moving. After controlling the vehicle transfer robot to move the required distance, control the vehicle transfer robot to stop moving. 5.3 If the environmental point cloud information in the longitudinal direction of the vehicle transfer robot obtained in step 5.1 does not contain vehicle information, then when the vehicle transfer robot reaches the end of the global vehicle placement path, the vehicle transfer robot control system will control the vehicle transfer robot to stop moving.

Images