An automatic parking method and system applied to a logistics park and a storage medium

Abstract

The application relates to an automatic parking method, system and storage medium applied to a logistics park, the method comprising the following steps: L1. A vehicle drives to a parking area of the logistics park, current position data information of the vehicle, target parking area data information and obstacle data information are acquired, a fast obstacle avoidance algorithm of a likelihood field is adopted to simulate a path of the vehicle, and path data information of the vehicle to the target parking area is obtained; L2. Based on the path data information of the vehicle to the target parking area, each piece of path data information is divided by setting a fixed step length, and a directed acyclic graph of each path is constructed, and directed acyclic graph group data information of the path is obtained. The application can not only realize real-time monitoring, but also has flexible automatic parking, does not need manual driving to a starting point, screen clicking to start automatic parking, reduces on-site manual work, and improves automatic loading and unloading of the park.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and in particular to an automatic parking method, system, and storage medium for use in logistics parks. Background Technology

[0002] With the continuous development of autonomous vehicles, according to incomplete statistics, there are currently more than 22,000 large-scale industrial parks and manufacturing parks nationwide, with a demand for more than 600,000 logistics vehicles. These parks require a large number of logistics drivers to complete short-distance, high-frequency bulk material transfers between warehouses, storage yards, and production lines within enclosed industrial areas, representing a labor-intensive industrial sector. Hardware, including LiDAR, cameras, millimeter-wave radar, and ultrasonic sensors, along with 5G communication equipment, and domestically produced controllers and self-developed autonomous driving programs, has enabled short-haul autonomous material transportation. This system possesses unmanned autonomous driving, automatic charging, and docking capabilities with unmanned warehousing systems, achieving a complete functional chain for continuous unmanned operation.

[0003] Currently, Automated Parking Systems (ADP) help drivers achieve unattended parking. First, turn on the car's power, press the system start button, and ensure the vehicle is in a safe parking position. Never activate the ADP while driving. The parking mode must be selected via the control panel or on-screen menu. The vehicle's hardware scans for parking spaces, and on-screen prompts guide the selection. After selection, move the vehicle to the appropriate starting point and press the system start button to begin parking. When automatic parking is complete, the system will emit an audible alert or display a parking completion message. At this point, the driver can turn off the system and shut down the vehicle. However, this system requires manual intervention to continuously select the appropriate starting point, power on, start the system, and select a suitable parking space. It does not eliminate manual labor and lacks automation, making it unsuitable for automated logistics connections within industrial parks, or for automating parking and loading / unloading processes. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the present invention provides an automatic parking method, system and storage medium for use in logistics parks. It can not only monitor in real time, but also make automatic parking more flexible. It eliminates the need for manual driving to the starting point and clicking on the screen to start automatic parking. Furthermore, it reduces on-site manpower and improves the automation of loading and unloading in the park.

[0005] To achieve the above and other related objectives, the present invention provides the following technical solution: An automated parking method for use in logistics parks, the method comprising: L1. When the vehicle drives to the parking area of ​​the logistics park, the current location data of the vehicle, the target parking area data and the obstacle data are obtained. The fast obstacle avoidance algorithm of the likelihood field is used to simulate the path of the vehicle and obtain the path data of the vehicle to the target parking area. L2. Based on the path data information from the vehicle to the target parking area, a fixed step size is set to divide the path data information into segments, and a directed acyclic graph of each path is constructed to obtain the directed acyclic graph group data information of the path. L3. Based on the directed acyclic graph group data information of the path, the improved Yorkson algorithm is used to perform topological sorting on the directed acyclic graph group of the path to obtain the optimal path data information; L4. Based on the optimal path data information, an improved pure geometric tracking algorithm is used to adjust and control the vehicle's steering angle and speed to complete parking and output the vehicle's control data information.

[0006] Furthermore, in step L1, the simulation of the vehicle's path using the likelihood field-based fast obstacle avoidance algorithm includes: L11. Based on the vehicle's current location data, target parking area data, and obstacle data, establish a vehicle obstacle avoidance probability function G. , Among them, X f X represents the matrix data of trajectory points from the vehicle's current location to the target parking area. s This is the drivable boundary matrix data information for the vehicle to reach the target parking area from its current position, where u is the collision probability density function between the vehicle and obstacles, and v is the probability distribution function between the target parking area and obstacles. L12. Based on the vehicle obstacle avoidance probability function G, perform probability analysis on the vehicle's drivable trajectory points to obtain obstacle avoidance probability data information for the vehicle trajectory points; L13. Based on the obstacle avoidance probability data of the vehicle trajectory points, a preset threshold is set. If the obstacle avoidance probability data of the vehicle trajectory points is less than the preset threshold, it is retained; if it is greater than the preset threshold, it is discarded, thus obtaining the vehicle drivable trajectory point data. L14. Based on the vehicle's drivable trajectory point data, a quintic spline function H is used. H = a0 + a1x + a2x 2 +ɑ3x 3 +ɑ4x 4 +ɑ5x 5 , The vehicle's driving path is fitted to obtain path data information from the vehicle to the target parking area.

[0007] Furthermore, in step L11, the collision probability density function u between the vehicle and the obstacle is, , Among them, X fThis is the trajectory point matrix data information of the vehicle from its current position to the target parking area, where η is the scale parameter of the vehicle's position, and X0 is the initial position parameter matrix of the vehicle.

[0008] Furthermore, in step L11, the probability distribution function v of the target parking area and the obstacle is, , Among them, X f X represents the matrix data of trajectory points from the vehicle's current location to the target parking area. s This provides the drivable boundary matrix data for the vehicle to reach the target parking area from its current location, where ω1, ω2, and ω3 represent X... f The parameter transformation matrix is ​​ω4 and ω5, which are the parameter transformation matrices of Xs, respectively.

[0009] Furthermore, in step L2, the construction of the directed acyclic graph for each path involves dividing the path data information into segments based on a fixed step size, obtaining the segmented path data information, and then constructing a directed acyclic graph for the segmented path data information according to a time series. The directed acyclic graph for each path includes edges and nodes, where nodes are the starting and ending points of the segmented path, and edges are the segments of the path.

[0010] Furthermore, in step L3, the topological sorting of the directed acyclic graph group of paths using the improved Yorkson algorithm includes: L31. Based on the directed acyclic graph group data information of the path, establish the cost function M of the directed acyclic graph of the path. , Where, r i Let λ be the length of the edge at the i-th node in the directed acyclic graph group of paths. i Let be the weight of the i-th node; L32. Based on the cost function M of the directed acyclic graph of the path, obtain the cost data information of each node in the directed acyclic graph group; L33. Based on the cost data information of each node in the directed acyclic graph group, compare and sort them to obtain the optimal path data information.

[0011] Furthermore, in step L33, the comparison and sorting involves selecting the node with the highest replacement value based on the cost data information of each node in the directed acyclic graph group.

[0012] Furthermore, in step L4, the adjustment and control of the vehicle's steering angle and speed using the improved pure geometry tracking algorithm includes: L41. Based on the optimal path data information, establish the vehicle's steering angle function N. , e(t) = h d sin(θt) Where t represents different time points, θ represents the tangent angle data of the optimal path of the vehicle at time t, and h d Let P be the path length at time t, P be the vehicle length, and e(t) be the vehicle lateral deviation. L42. Based on the vehicle's steering angle function N, obtain the vehicle's steering angle data information; L43. Based on the vehicle's steering angle data, the vehicle's speed is adjusted to obtain the vehicle's control data.

[0013] To achieve the above and other related objectives, the present invention also provides a system for implementing any of the automatic parking methods applied to logistics parks, including a parking space monitoring system, a cloud platform system, and a vehicle automatic driving system. The parking space monitoring system pushes the parking space status to the cloud platform system in real time via HTTP protocol through a server backend, including device ID, parking space number, installation location, latitude and longitude, parking space status, and timestamp. The cloud platform system connects to the upstream logistics park production system to perform global scheduling and full automation of vehicle tailgate opening, loading, driving, parking, unloading, and tailgate closing. The cloud platform Web API interface uses HTTP as the network protocol for data transmission, and JSON as the data format for data exchange. The vehicle autonomous driving system calculates the optimal front wheel steering angle and speed based on vehicle status information and target parking space information to achieve automatic parking alignment. Based on the calculated front wheel steering angle and speed, it controls the vehicle to automatically perform parking operations, achieving accurate alignment with the target parking space.

[0014] To achieve the above and other related objectives, the present invention also provides a computer-readable storage medium storing a computer program programmed or configured to perform any of the described automated parking methods for logistics parks.

[0015] The present invention has the following positive effects: 1. This invention uses a fast obstacle avoidance algorithm based on likelihood fields to simulate the vehicle's path, obtaining path data information from the vehicle to the target parking area. By setting a fixed step size to divide the data information of each path, and constructing a directed acyclic graph for each path, this invention not only enables global analysis of the vehicle's driving path, but also ensures that the vehicle can complete the parking task safely and efficiently.

[0016] 2. This invention employs an improved Yorkson algorithm to perform topological sorting on the directed acyclic graph of the path, and combines this with an improved pure geometric tracking algorithm to adjust and control the vehicle's steering angle and speed to complete parking. This not only makes automatic parking more flexible, eliminating the need for manual driving to the starting point and clicking on the screen to begin automatic parking, but also reduces on-site manpower and improves the automation of loading and unloading in the park. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the system framework of the present invention; Figure 3 This is a schematic diagram illustrating the process of simulating the vehicle's path using the likelihood field-based fast obstacle avoidance algorithm of this invention. Detailed Implementation

[0018] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0019] Example 1: As Figure 1 As shown, an automated parking method applied to a logistics park includes: L1. When the vehicle drives to the parking area of ​​the logistics park, the current location data of the vehicle, the target parking area data and the obstacle data are obtained. The fast obstacle avoidance algorithm of the likelihood field is used to simulate the path of the vehicle and obtain the path data of the vehicle to the target parking area. L2. Based on the path data information from the vehicle to the target parking area, a fixed step size is set to divide the path data information into segments, and a directed acyclic graph of each path is constructed to obtain the directed acyclic graph group data information of the path. L3. Based on the directed acyclic graph group data information of the path, the improved Yorkson algorithm is used to perform topological sorting on the directed acyclic graph group of the path to obtain the optimal path data information; L4. Based on the optimal path data information, an improved pure geometric tracking algorithm is used to adjust and control the vehicle's steering angle and speed to complete parking and output the vehicle's control data information.

[0020] In this embodiment, as Figure 3 As shown, in step L1, the simulation of the vehicle's path using the likelihood field-based fast obstacle avoidance algorithm includes: L11. Based on the vehicle's current location data, target parking area data, and obstacle data, establish a vehicle obstacle avoidance probability function G. , Among them, X f X represents the matrix data of trajectory points from the vehicle's current location to the target parking area. s This is the drivable boundary matrix data information for the vehicle to reach the target parking area from its current position, where u is the collision probability density function between the vehicle and obstacles, and v is the probability distribution function between the target parking area and obstacles. L12. Based on the vehicle obstacle avoidance probability function G, perform probability analysis on the vehicle's drivable trajectory points to obtain obstacle avoidance probability data information for the vehicle trajectory points; L13. Based on the obstacle avoidance probability data of the vehicle trajectory points, a preset threshold is set. If the obstacle avoidance probability data of the vehicle trajectory points is less than the preset threshold, it is retained; if it is greater than the preset threshold, it is discarded, thus obtaining the vehicle drivable trajectory point data. L14. Based on the vehicle's drivable trajectory point data, a quintic spline function H is used. H = a0 + a1x + a2x 2 +ɑ3x 3 +ɑ4x 4 +ɑ5x 5 , The vehicle's driving path is fitted to obtain path data information from the vehicle to the target parking area.

[0021] In this embodiment, in step L11, the collision probability density function u between the vehicle and the obstacle is, , Among them, X f This is the trajectory point matrix data information of the vehicle from its current position to the target parking area, where η is the scale parameter of the vehicle's position, and X0 is the initial position parameter matrix of the vehicle.

[0022] In this embodiment, in step L11, the probability distribution function v of the target parking area and the obstacle is, , Among them, X f X represents the matrix data of trajectory points from the vehicle's current location to the target parking area. s This provides the drivable boundary matrix data for the vehicle to reach the target parking area from its current location, where ω1, ω2, and ω3 represent X... f The parameter transformation matrix is ​​ω4 and ω5, which are the parameter transformation matrices of Xs, respectively.

[0023] Example 2: Based on the automatic parking method applied to a logistics park in Example 1, the present invention will be further explained and described below.

[0024] like Figure 1 As shown, an automated parking method applied to a logistics park includes: L1. When the vehicle drives to the parking area of ​​the logistics park, the current location data of the vehicle, the target parking area data and the obstacle data are obtained. The fast obstacle avoidance algorithm of the likelihood field is used to simulate the path of the vehicle and obtain the path data of the vehicle to the target parking area. L2. Based on the path data information from the vehicle to the target parking area, a fixed step size is set to divide the path data information into segments, and a directed acyclic graph of each path is constructed to obtain the directed acyclic graph group data information of the path. L3. Based on the directed acyclic graph group data information of the path, the improved Yorkson algorithm is used to perform topological sorting on the directed acyclic graph group of the path to obtain the optimal path data information; L4. Based on the optimal path data information, an improved pure geometric tracking algorithm is used to adjust and control the vehicle's steering angle and speed to complete parking and output the vehicle's control data information.

[0025] In this embodiment, in step L2, constructing the directed acyclic graph for each path involves dividing the path data information into segments based on a fixed step size, obtaining the segmented path data information, and then constructing the directed acyclic graph for the segmented path data information according to a time series. The directed acyclic graph for each path includes edges and nodes, where nodes are the starting and ending points of the segmented path, and edges are the segments of the path.

[0026] In this embodiment, step L3, which involves performing topological sorting on the directed acyclic graph group of paths using the improved Yorkson algorithm, includes: L31. Based on the directed acyclic graph group data information of the path, establish the cost function M of the directed acyclic graph of the path. , Where, r i Let λ be the length of the edge at the i-th node in the directed acyclic graph group of paths. i Let be the weight of the i-th node; L32. Based on the cost function M of the directed acyclic graph of the path, obtain the cost data information of each node in the directed acyclic graph group; L33. Based on the cost data information of each node in the directed acyclic graph group, compare and sort them to obtain the optimal path data information.

[0027] In this embodiment, in step L33, the comparison and sorting is to select the node with the highest replacement value based on the cost data information of each node in the directed acyclic graph group.

[0028] In this embodiment, step L4, which involves adjusting and controlling the vehicle's steering angle and speed using an improved pure geometry tracking algorithm, includes: L41. Based on the optimal path data information, establish the vehicle's steering angle function N. , e(t) = h d sin(θt) Where t represents different time points, θ represents the tangent angle data of the optimal path of the vehicle at time t, and h d Let P be the path length at time t, P be the vehicle length, and e(t) be the vehicle lateral deviation. L42. Based on the vehicle's steering angle function N, obtain the vehicle's steering angle data information; L43. Based on the vehicle's steering angle data, the vehicle's speed is adjusted to obtain the vehicle's control data.

[0029] like Figure 2 As shown, the present invention provides a system for implementing any of the automatic parking methods applied to logistics parks, including a parking space monitoring system, a cloud platform system, and a vehicle automatic driving system. The parking space monitoring system pushes the parking space status to the cloud platform system in real time via HTTP protocol through a server backend, including device ID, parking space number, installation location, latitude and longitude, parking space status, and timestamp. The cloud platform system connects to the upstream logistics park production system to perform global scheduling and full automation of vehicle tailgate opening, loading, driving, parking, unloading, and tailgate closing. The cloud platform Web API interface uses HTTP as the network protocol for data transmission, and JSON as the data format for data exchange. The vehicle autonomous driving system calculates the optimal front wheel steering angle and speed based on vehicle status information and target parking space information to achieve automatic parking alignment. Based on the calculated front wheel steering angle and speed, it controls the vehicle to automatically perform parking operations, achieving accurate alignment with the target parking space.

[0030] The present invention also provides a computer-readable storage medium storing a computer program programmed or configured to perform any of the described automated parking methods for logistics parks.

[0031] Any references to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0032] In summary, this invention not only enables real-time monitoring and more flexible automatic parking, eliminating the need for manual driving to the starting point and simply clicking on the screen to begin automatic parking, but also reduces on-site manpower and improves automated loading and unloading in the park.

[0033] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. An automated parking method for use in logistics parks, characterized in that, The method includes: L1. When the vehicle drives to the parking area of ​​the logistics park, the current location data of the vehicle, the target parking area data and the obstacle data are obtained. The fast obstacle avoidance algorithm of the likelihood field is used to simulate the path of the vehicle and obtain the path data of the vehicle to the target parking area. L2. Based on the path data information from the vehicle to the target parking area, a fixed step size is set to divide the path data information into segments, and a directed acyclic graph of each path is constructed to obtain the directed acyclic graph group data information of the path. L3. Based on the directed acyclic graph group data information of the path, the improved Yorkson algorithm is used to perform topological sorting on the directed acyclic graph group of the path to obtain the optimal path data information; L4. Based on the optimal path data information, an improved pure geometric tracking algorithm is used to adjust and control the vehicle's steering angle and speed to complete parking and output the vehicle's control data information. In step L1, the simulation of the vehicle's path using the fast obstacle avoidance algorithm with a likelihood field includes: L11. Based on the vehicle's current location data, target parking area data, and obstacle data, establish a vehicle obstacle avoidance probability function G. ， Among them, X f X represents the matrix data of trajectory points from the vehicle's current location to the target parking area. s This is the drivable boundary matrix data information for the vehicle to reach the target parking area from its current position, where u is the collision probability density function between the vehicle and obstacles, and v is the probability distribution function between the target parking area and obstacles. L12. Based on the vehicle obstacle avoidance probability function G, perform probability analysis on the vehicle's drivable trajectory points to obtain obstacle avoidance probability data information for the vehicle trajectory points; L13. Based on the obstacle avoidance probability data of the vehicle trajectory points, a preset threshold is set. If the obstacle avoidance probability data of the vehicle trajectory points is less than the preset threshold, it is retained; if it is greater than the preset threshold, it is discarded, thus obtaining the vehicle drivable trajectory point data. L14. Based on the vehicle's drivable trajectory point data, a quintic spline function H is used. H=ɑ0+ɑ1x+ɑ2x 2 +ɑ3x 3 +ɑ4x 4 +ɑ5x 5 , The vehicle's driving path is fitted to obtain path data information from the vehicle to the target parking area; In step L3, the topological sorting of the directed acyclic graph group of paths using the improved Yorkson algorithm includes: L31. Based on the directed acyclic graph group data information of the path, establish the cost function M of the directed acyclic graph of the path. ， Where, r i Let λ be the length of the edge at the i-th node in the directed acyclic graph group of paths. i Let be the weight of the i-th node; L32. Based on the cost function M of the directed acyclic graph of the path, obtain the cost data information of each node in the directed acyclic graph group; L33. Based on the cost data information of each node in the directed acyclic graph group, compare and sort them to obtain the optimal path data information; In step L4, the adjustment and control of the vehicle's steering angle and speed using the improved pure geometry tracking algorithm includes: L41. Based on the optimal path data information, establish the vehicle's steering angle function N. ， e(t)=h d sin(θt) Where t represents different time points, θ represents the tangent angle data of the optimal path of the vehicle at time t, and h d Let P be the path length at time t, P be the vehicle length, and e(t) be the vehicle lateral deviation. L42. Based on the vehicle's steering angle function N, obtain the vehicle's steering angle data information; L43. Based on the vehicle's steering angle data, the vehicle's speed is adjusted to obtain the vehicle's control data.

2. The automatic parking method for logistics parks according to claim 1, characterized in that, In step L11, the collision probability density function u between the vehicle and the obstacle is, ， Among them, X f This is the trajectory point matrix data information of the vehicle from its current position to the target parking area, where η is the scale parameter of the vehicle's position, and X0 is the initial position parameter matrix of the vehicle.

3. The automatic parking method for logistics parks according to claim 2, characterized in that, In step L11, the probability distribution function v of the target parking area and obstacles is, ， Among them, X f X represents the matrix data of trajectory points from the vehicle's current location to the target parking area. s This provides the drivable boundary matrix data for the vehicle to reach the target parking area from its current location, where ω1, ω2, and ω3 represent X... f The parameter transformation matrix is ​​ω4 and ω5, which are the parameter transformation matrices of Xs, respectively.

4. The automatic parking method for logistics parks according to claim 1, characterized in that, In step L2, constructing the directed acyclic graph for each path involves dividing the path data information into segments based on a fixed step size, obtaining the segmented path data information, and then constructing a directed acyclic graph for the segmented path data information according to a time series. The directed acyclic graph for each path includes edges and nodes, where nodes are the starting and ending points of the segmented path, and edges are the segments of the path.

5. The automatic parking method for logistics parks according to claim 1, characterized in that, In step L33, the comparison and sorting involves selecting the node with the highest replacement value based on the cost data information of each node in the directed acyclic graph group.

6. A system for implementing the automated parking method for logistics parks as described in any one of claims 1-5, comprising a parking space monitoring system, a cloud platform system, and a vehicle automated driving system, characterized in that: The parking space monitoring system pushes the parking space status to the cloud platform system in real time via HTTP protocol through the server backend, including device ID, parking space number, installation location, latitude and longitude, parking space status and timestamp; The cloud platform system connects to the upstream logistics park production system to perform global scheduling and full automation of vehicle tailgate opening, loading, driving, parking, unloading, and tailgate closing. The cloud platform Web API interface uses HTTP as the network protocol for data transmission, and JSON as the data format for data exchange. The vehicle autonomous driving system calculates the optimal front wheel steering angle and speed based on vehicle status information and target parking space information to achieve automatic parking alignment. Based on the calculated front wheel steering angle and speed, it controls the vehicle to automatically perform parking operations, achieving accurate alignment with the target parking space.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is programmed or configured to perform the automated parking method for a logistics park as described in any one of claims 1 to 5.

Images