A parking path planning method and device

Abstract

The application relates to a parking path planning method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: firstly, acquiring a no-collision pose point in a preset area to which a target parking pose point belongs; wherein the no-collision pose point is a pose point having a no-collision path with the target parking pose point. Then, it is determined whether a no-collision path exists from an intermediate pose point to the no-collision pose point. If a no-collision path exists, a final parking path is generated according to a no-collision path from a current pose point to the intermediate pose point, a no-collision path from the intermediate pose point to the no-collision pose point and a no-collision path from the no-collision pose point to the target parking pose point. Through the method, the efficiency of parking path planning is improved in a complex scene of obstacles near the target parking pose point.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of path planning, in particular to a parking path planning method and device. BACKGROUND

[0002] With the development of automatic driving technology, path planning technology appears, and the most widely used path planning technology is the path planning strategy based on A* algorithm. The path planning strategy based on hybrid A* algorithm further improves the path planning strategy based on A* algorithm, and the planned path is more consistent with the kinematics of the vehicle. In the process of planning the path based on the hybrid A* algorithm, the search pose points are gradually expanded from the current pose point to the front, forming a travel path.

[0003] However, in actual application, in the scenario of parking, there are usually more obstacles near the parking point. When the path planning strategy based on the hybrid A* algorithm expands the search pose points near the parking point, the search difficulty is greatly improved, resulting in low search efficiency near the parking point. SUMMARY

[0004] Therefore, it is necessary to provide a parking path planning method and device, computer equipment, computer readable storage medium and computer program product capable of improving the parking path planning efficiency to solve the above technical problems.

[0005] In a first aspect, the present application provides a parking path planning method. The method comprises:

[0006] obtaining a collision-free pose point in a preset area to which a target parking pose point belongs; the collision-free pose point is a pose point having a collision-free path with the target parking pose point;

[0007] determining whether there is a collision-free path from an intermediate pose point to the collision-free pose point; the intermediate pose point is a pose point between the current pose point and the target parking pose point determined according to a preset path search strategy;

[0008] if there is, generating a parking path according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point.

[0009] In one of the embodiments, the method further comprises:

[0010] if there is no collision-free path from the intermediate pose point to the collision-free pose point, the intermediate pose point is determined again according to the path search strategy, and the step of determining whether there is a collision-free path from the intermediate pose point to the collision-free pose point is performed.

[0011] In one of the embodiments, the method further comprises:

[0012] determining a plurality of first candidate pose points in the preset region to which the target parking pose point belongs, according to a preset pose point selection strategy;

[0013] for each first candidate pose point, if there is a collision-free path from the first candidate pose point to the target parking pose point, the first candidate pose point is determined as a collision-free pose point.

[0014] In one of the embodiments, the method further comprises:

[0015] determining a plurality of position points in the preset region according to a preset position selection strategy;

[0016] determining a pose point corresponding to each position point as a first candidate pose point according to a preset pose angle.

[0017] In one of the embodiments, the method further comprises:

[0018] determining a plurality of first candidate pose points in the preset region to which the target parking pose point belongs, according to a preset pose point selection strategy;

[0019] for each first candidate pose point, if there is a collision-free path from the first candidate pose point to the target parking pose point, the first candidate pose point is determined as a second candidate pose point;

[0020] for each second candidate pose point, determining a driving cost value of the second candidate pose point driving to the target parking pose point;

[0021] determining the second candidate pose point with the minimum driving cost value as a collision-free pose point.

[0022] In one of the embodiments, the method further comprises:

[0023] For each second candidate pose point, obtain a driving cost parameter of the second candidate pose point, wherein the driving cost parameter comprises one or more of a Voronoi potential value of the second candidate pose point, a gradient value of the Voronoi potential along a pose direction of the second candidate pose point, a Voronoi potential integral value on a path from the second candidate pose point to the target parking pose point, and a pose direction difference value between the intermediate pose point and the second candidate pose point.

[0024] According to the obtained driving cost parameter of the second candidate pose point, determine a driving cost value of the second candidate pose point.

[0025] In a second aspect, the present application also provides a parking path planning device. The device comprises:

[0026] An obtaining module is configured to obtain a collision-free pose point in a preset region to which a target parking pose point belongs; the collision-free pose point is a pose point having a collision-free path with the target parking pose point;

[0027] A determining module is configured to determine whether a collision-free path exists from an intermediate pose point to the collision-free pose point; the intermediate pose point is a pose point between a current pose point and the target parking pose point, which is determined according to a preset path search strategy;

[0028] A generating module is configured to, if a collision-free path exists, generate a parking path according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point.

[0029] In a third aspect, the present application also provides a computer device. The computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0030] Obtain a collision-free pose point in a preset region to which a target parking pose point belongs; the collision-free pose point is a pose point having a collision-free path with the target parking pose point;

[0031] Determine whether a collision-free path exists from an intermediate pose point to the collision-free pose point; the intermediate pose point is a pose point between a current pose point and the target parking pose point, which is determined according to a preset path search strategy;

[0032] If a collision-free path exists, generate a parking path according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point.

[0033] In a fourth aspect, the present application also provides a computer readable storage medium. The computer readable storage medium has a computer program stored thereon, and the computer program, when executed by a processor, implements the following steps:

[0034] obtaining a collision-free pose point in a preset area to which the target parking pose point belongs; the collision-free pose point is a pose point having a collision-free path with the target parking pose point;

[0035] determining whether there is a collision-free path from an intermediate pose point to the collision-free pose point; the intermediate pose point is a pose point between the current pose point and the target parking pose point, which is determined according to a preset path search strategy;

[0036] if there is, generating a parking path according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point.

[0037] In a fifth aspect, the present application also provides a computer program product. The computer program product comprises a computer program, and the computer program, when executed by a processor, implements the following steps:

[0038] obtaining a collision-free pose point in a preset area to which the target parking pose point belongs; the collision-free pose point is a pose point having a collision-free path with the target parking pose point;

[0039] determining whether there is a collision-free path from an intermediate pose point to the collision-free pose point; the intermediate pose point is a pose point between the current pose point and the target parking pose point, which is determined according to a preset path search strategy;

[0040] if there is, generating a parking path according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point.

[0041] The parking path method, device, computer device, storage medium and computer program product described above, in the scenario of parking path planning, effectively accelerate the path planning efficiency near the parking pose point, because the last segment of the parking path is planned in advance. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 a basic flowchart based on a hybrid A* path planning strategy according to an embodiment of the present application;

[0043] Figure 2A flowchart of a parking path planning method according to an embodiment of the present application is shown in the figure;

[0044] Figure 3 A schematic diagram of each pose point in a preset area to which a target pose point according to an embodiment of the present application belongs is shown in the figure;

[0045] Figure 4 A flowchart of a parking path planning method according to another embodiment of the present application is shown in the figure;

[0046] Figure 5 A structural block diagram of a parking path planning device according to an embodiment of the present application is shown in the figure;

[0047] Figure 6 An internal structure diagram of a computer device in an embodiment. DETAILED DESCRIPTION

[0048] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0049] In the related art, the path planning strategy based on the hybrid A* algorithm considers the actual motion constraints of the object on the basis of the A*-based algorithm path planning strategy. The basic flow of the hybrid A* algorithm path planning strategy is as shown in the figure. Figure 1

[0050] 1. First, determine the current pose point and the target pose point of the path planning of the vehicle driving, and initialize the open set as the current pose point and the closed set as empty.

[0051] 2. Determine whether the open set is empty. If it is empty, it means that the path search fails; if it is not empty, continue to execute step 3.

[0052] 3. Determine the total cost f of each pose point in the open set, and select the pose point with the minimum cost in the open set as the intermediate pose point, and add the intermediate pose point to the closed set. The total cost f of each pose point includes two parts, the motion cost item g and the heuristic cost item h. The motion cost item g is used to represent the driving value from the current pose point to the pose point, and the heuristic cost h is used to represent the heuristic value from the pose point to the target pose point.

[0053] 4. Detect whether there is a collision-free path between the intermediate pose point and the target pose point. If there is, the path from the current pose point to the intermediate pose point and the collision-free path from the intermediate pose point to the target pose point are taken as the final planned path; if there is not, continue to execute step 5.

[0054] ​5. Selecting a neighboring pose point of the intermediate pose point according to the kinematic formula.

[0055] 6. Determining whether there is a collision-free path from the neighboring pose point to the intermediate pose point. If yes, continue to step 7; if no, return to step 5.

[0056] 7. Determining whether the neighboring pose point is in the closed set. If yes, return to step 5; if no, continue to step 8.

[0057] 8. Calculating the motion cost g from the current pose point to the neighboring pose point according to the motion cost from the current pose point to the intermediate pose point and the motion cost from the intermediate pose point to the neighboring pose point.

[0058] 9. Determining whether the neighboring pose point is in the open set. If yes, continue to step 10; if no, continue to step 12.

[0059] 10. Determining the size of the motion cost g of the neighboring pose point and the motion cost g of the neighboring pose point saved in the open set. If the motion cost g of the neighboring pose point is not less than the motion cost g of the neighboring pose point saved in the open set, return to step 2 directly. If the motion cost g of the neighboring pose point is less than the motion cost g of the neighboring pose point saved in the open set, continue to step 11.

[0060] 11. Updating the motion cost g, the parent node and the total cost f of the neighboring pose point in the open set, and returning to step 2.

[0061] 12. Calculating the heuristic cost h of the neighboring pose point, and calculating the total cost f of the neighboring pose point according to the motion cost g and the heuristic cost h of the neighboring pose point, recording the parent node as the intermediate pose point, and adding the neighboring pose point to the open set. Return to step 2.

[0062] According to the basic flow of the above-mentioned hybrid A* algorithm path planning strategy, it can be known that, when approaching the target pose point, in order to connect to the target pose as low as possible and improve the speed of the search, the hybrid A* algorithm attempts to find a path without considering obstacles. Therefore, when there are many obstacles near the target pose point, especially the U-shaped obstacle in the parking scene, the difficulty of searching for a neighboring pose point near the target pose point is greatly improved, and the efficiency is greatly reduced.

[0063] Based on this, the application shows a path planning method for complex obstacles near the target pose point based on the path planning strategy based on the hybrid A* algorithm, that is, a parking path planning method. First, the collision-free pose points in the preset area to which the target parking pose point belongs are obtained. The collision-free pose point is a pose point that has a collision-free path with the target parking pose point. Then it is determined whether there is a collision-free path from the intermediate pose point to the collision-free pose point. If there is, the final parking path is generated according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point.

[0064] By the method of the embodiment of the application, the last parking path planning with the most obstacles is completed according to the collision-free pose points near the target parking pose point. Therefore, when the parking path is planned, only the path from the current pose point to the starting pose point (collision-free pose point) of the last parking path needs to be planned, and the obstacles near the collision-free pose point are much less than the obstacles near the target parking pose point. Therefore, the overall efficiency of the parking path planning is improved.

[0065] First, the parking path planning method shown in the application is described in detail. The parking path planning method provided by the embodiment of the application can be applied to, but is not limited to, personal computers, notebook computers, smart phones, tablet computers, Internet of Things devices, portable wearable devices, independent servers, or server clusters composed of multiple servers.

[0066] As shown in Figure 2 The application provides a flowchart of a parking path planning method according to an embodiment, which includes the following steps:

[0067] Step 201, obtaining the collision-free pose points in the preset area to which the target parking pose point belongs.

[0068] The collision-free pose points are located in the preset area to which the target parking pose point belongs, and the collision-free pose points have a collision-free path with the target parking pose point. The preset area to which the target parking pose point belongs can be a region with a fixed length and width, such as a region with a length of 11 meters and a width of 5 meters, and the target parking pose point is in the center of the region. The device can also divide a region with a fixed length and width according to the obstacles near the target parking pose point, so that the target parking pose point is close to one side of the region, and the obstacles on this side are the most.

[0069] In one embodiment, each pose point includes three information, the horizontal coordinate in the map, the vertical coordinate in the map, and the orientation of the vehicle, that is, the same horizontal coordinate and vertical coordinate can have multiple pose points. For example, a pose point can include three dimensions (x, y, θ), representing the position (x, y) in the plane xOy coordinate system and the vehicle orientation angle θ.

[0070] In one embodiment, the device stores map data containing obstacle information, and explicitly the current pose point and the target parking pose point of the parking path planning, and the basic parameters of the actual driving vehicle for the path planning, such as the length and width of the actual driving vehicle, and the kinematics related parameters of the vehicle, such as the wheelbase, the distance between the rear axle and the front end, and the minimum turning radius. Therefore, when determining whether there is a collision-free path between two pose points, the device can generate a driving path that meets the kinematics of the vehicle according to the preset path generation strategy and the kinematics parameters of the vehicle (such as the minimum turning radius), and then determine whether the vehicle will collide along the generated path according to the basic parameters of the vehicle, such as the length and width of the vehicle. If there is a collision, the device determines that there is no collision-free path between the two pose points; if there is no collision, the device determines that there is a collision-free path between the two pose points.

[0071] The preset path generation strategy can be an RS curve path generation strategy or a Dubins curve path generation strategy. RS curve and Dubins curve are similar, both of which are circular + straight line paths, and the difference is that RS curve allows the vehicle to reverse, and Dubins curve only allows the vehicle to move forward.

[0072] As Figure 3 shown in the schematic diagram of each pose point in the preset area to which the target parking pose point belongs, the position of the solid hexagonal star represents the position of the target parking pose point, the arrow to which the solid hexagonal star belongs represents the orientation of the target parking pose point, the dashed line near the target parking pose point is the obstacle of the target parking pose point, and the position of the solid origin represents the position of a pose point, and the arrow to which the solid origin belongs represents the orientation of the pose point. The pose point circled by the circle is a collision-free pose point, and the line between the collision-free pose point and the target parking pose point represents a collision-free path from the collision-free pose point to the target parking pose point. There is no collision-free path from other pose points to the target parking pose point.

[0073] Step 203, determining whether there is a collision-free path from the intermediate pose point to the collision-free pose point.

[0074] The intermediate pose point is determined according to a preset path search strategy between the current pose point and the target parking pose point. When the pose point is expanded according to the preset path search strategy, the intermediate pose point is constantly updated according to the path search situation. Taking the preset path search strategy as the path planning based on the mixed A* algorithm as an example, each time the loop is executed to step 3, a pose point with the minimum total cost f is selected from the open set as the intermediate pose point until there is a collision-free path from the intermediate pose point to the target pose point. The difference lies in that the method of the application is until there is a collision-free path from the intermediate pose point to the collision-free pose point.

[0075] In one embodiment, when determining whether there is a collision-free path from the intermediate pose point to the collision-free pose point, the device first determines whether the path from the intermediate pose point to the collision-free pose point conforms to the vehicle kinematics without considering the obstacle information in the map. If it is determined that there is a path from the intermediate pose point to the collision-free pose point that conforms to the vehicle kinematics, the device further determines whether the path exists obstacles according to the obstacle information in the map data, and determines whether a collision will occur in actual driving according to the basic information of the actual driving vehicle. The device determines that there is a collision-free path from the intermediate pose point to the collision-free pose point in the case that the path will not collide.

[0076] In the case of not considering the obstacle information in the map, the determination of whether the path from the intermediate pose point to the collision-free pose point conforms to the vehicle kinematics can be a preset path generation strategy, such as RS curve path generation algorithm or Dubins curve path generation algorithm. RS curve and Dubins curve are similar, both of which are circular + straight line paths, and the difference lies in that RS curve allows the vehicle to reverse, and Dubins curve only allows the vehicle to move forward. If the preset path generation strategy can generate a path from the intermediate pose point to the collision-free pose point that conforms to the vehicle kinematics, then the path from the intermediate pose point to the collision-free pose point conforms to the vehicle kinematics. If the preset path generation strategy cannot generate a path from the intermediate pose point to the collision-free pose point that conforms to the vehicle kinematics, then the path from the intermediate pose point to the collision-free pose point does not conform to the vehicle kinematics.

[0077] Step 205, if there is, a parking path is generated according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point.

[0078] In one embodiment, the device plans the parking path from the current pose point to the target parking pose point, which includes three parts. The first part is to plan the path from the current pose point to the intermediate pose point according to the preset path search strategy. The second part is to generate the collision-free path from the intermediate pose point to the collision-free pose point according to the preset path generation strategy. The third part is to generate the collision-free path from the collision-free pose point to the target parking pose point according to the preset path generation strategy. The device combines the paths of the three parts to generate the planned parking path from the current pose point to the target parking pose point.

[0079] For example, the current pose point is A, the intermediate pose point is D, the collision-free pose point is X, and the target parking pose point is Y. The device plans the path from the current pose point A to the intermediate pose point D as A-B-C-D according to the preset path search strategy, determines that there is a collision-free path D-X from the intermediate pose point D to the collision-free pose point X, and that there is a collision-free path X-Y from the collision-free pose point X to the target parking pose point Y. Then the device generates the path A-B-C-D-X-Y.

[0080] In the process of planning the path from the current pose point to the intermediate pose point according to the preset path search strategy, the device can use the path planning strategy based on the hybrid A* algorithm or the path planning strategy based on the A* algorithm, which is not limited in the present application.

[0081] According to the present embodiment, the last part of the parking path, i.e., the third part of the parking path, is planned separately, and the planning difficulty of the second part of the parking path is reduced, thereby accelerating the overall efficiency of the parking path planning, especially the planning efficiency of the second and third parts of the parking path.

[0082] In one embodiment, there can be no collision-free path from the intermediate pose point to the collision-free pose point. In this case, the above method can also include:

[0083] In step 207, if there is no collision-free path from the intermediate pose point to the collision-free pose point, the device re-determines the intermediate pose point according to the path search strategy and executes step 203.

[0084] In one embodiment, if there is no collision-free path from the intermediate pose point to the collision-free pose point, the device can determine that it is still necessary to continue searching the path forward according to the preset path search strategy, expand a new pose point, i.e., update the intermediate pose point. Then, according to the updated intermediate pose point, it is determined whether there is a collision-free path from the intermediate pose point to the collision-free pose point.

[0085] In one embodiment, taking the path planning strategy based on the hybrid A* algorithm as an example, if the device determines that there is no collision-free path from the intermediate pose point to the collision-free pose point, it means that more pose points need to be expanded according to the preset path search strategy, and then the pose point with the minimum total cost f is selected as the intermediate pose point according to the total cost f of each pose point.

[0086] In the related art, the path planning strategy based on the hybrid A* algorithm may only expand to the pose point with the collision-free path to the target parking pose point when the target parking pose point is reached due to the many obstacles near the target parking pose point, so the device needs to update the intermediate pose point many times to expand the intermediate pose point to the pose point with the collision-free path to the target parking pose point.

[0087] In this embodiment, since there are relatively fewer obstacles near the collision-free pose point, the pose point with the collision-free path to the collision-free pose point can be expanded from a very far pose point, so the device does not need to repeatedly perform step 207 many times to expand the intermediate pose point to the pose point with the collision-free path to the collision-free pose point. This greatly improves the efficiency of path planning in a scene with complex obstacles near the target parking pose point.

[0088] In one embodiment, when determining the collision-free pose point near the target parking pose point, the candidate pose point selection strategy can be used to determine a plurality of first candidate pose points in a preset region to which the target parking pose point belongs, and then further filter the pose points that meet the condition of "having a collision-free path to the target parking pose point" as the collision-free pose points.

[0089] Therefore, in one embodiment, step 201 can specifically include:

[0090] Step 201a: determining a plurality of first candidate pose points in a preset region around the target parking pose point according to a preset pose point selection strategy.

[0091] The preset candidate pose point selection strategy is not limited in the present application, and can select a plurality of first candidate pose points that meet the condition of "being in the preset region around the target parking pose point".

[0092] In one embodiment, the preset candidate pose point selection strategy can be to randomly select a plurality of first candidate pose points in the preset region to which the target parking pose point belongs. The candidate pose point selection strategy can also be to obtain a plurality of first candidate pose points that are uniformly distributed in the preset region to which the target parking pose point belongs according to the length direction step and the width direction step.

[0093] Step 201b, for each first candidate pose point, if there is a collision-free path between the first candidate pose point and the target parking pose point, the first candidate pose point is taken as a collision-free pose point.

[0094] In one embodiment, after the device determines the plurality of first candidate pose points, for each first candidate pose point, it can directly determine whether there is a collision-free path between the first candidate pose point and the target parking pose point. If it is determined that there is a collision-free path between the first candidate pose point and the target parking pose point, the device can take the first candidate pose point as a collision-free pose point. If it is determined that there is no collision-free path between the first candidate pose point and the target parking pose point, the device directly discards the first candidate pose point.

[0095] In one embodiment, after the device determines the plurality of first candidate pose points, for each first candidate pose point, the device can further determine whether there is an obstacle at the position of the first candidate pose point and whether the vehicle will collide at the first candidate pose point according to the stored map data containing obstacle information, so as to filter out unnecessary first candidate pose points for determining whether there is a collision-free path. If it is determined that there is an obstacle at the position of the first candidate pose point or the vehicle will collide at the first candidate pose point, it means that the first candidate pose point is necessarily a pose point where the vehicle cannot stay, and the device directly discards the first candidate pose point, so as to reduce the consumption of computing resources for determining whether there is a collision-free path for the first candidate pose point. If it is determined that there is no obstacle at the position of the first candidate pose point and the vehicle will not collide at the first candidate pose point, the device further determines whether there is a collision-free path between the first candidate pose point and the target parking pose point. If it is determined that there is a collision-free path between the first candidate pose point and the target parking pose point, the device can take the first candidate pose point as a collision-free pose point. If it is determined that there is no collision-free path between the first candidate pose point and the target parking pose point, the device directly discards the first candidate pose point.

[0096] In this embodiment, each first candidate pose point near the target parking pose point is selected by the preset candidate pose point selection strategy, and then a collision-free pose point is further selected from the first candidate pose points.

[0097] In one embodiment, the preset candidate pose point selection strategy is to obtain a plurality of first candidate pose points distributed uniformly in a preset region where the target parking pose point is located according to the length direction step and the width direction step. At this time, the above step 201a specifically includes:

[0098] Step a1, determining a plurality of position points in a preset region where the target parking pose point is located based on a preset position selection strategy.

[0099] In one embodiment, a length direction step size and a width direction step size can be defined, and (pre-set area length / length direction step size + 1) * (pre-set area width / width direction step size + 1) position points can be determined in the pre-set area to which the target parking pose point belongs. For example, if the pre-set area to which the target parking pose point belongs has a length of 10 m and a width of 8 m, and the length direction step size and the width direction step size are both 1 m, then 11 * 9 = 99 pose points can be determined.

[0100] In step a2, based on the pre-set pose angle, a pose point corresponding to each position point is determined as a first candidate pose point.

[0101] In one embodiment, each position point has several directions, and an angle direction step size can be defined to obtain (2π / angle direction step size) orientation angles. The device can obtain (2π / angle direction step size) pose points at the position point according to the determined 0-degree orientation angle.

[0102] In one embodiment, assuming that a pre-set area determined near the target parking pose point is a region with a length of W and a width of H, and θ ∈ (-π, π] represents the definition domain of the orientation angle, the length direction step size, the width direction step size, and the angle direction step size in the three-dimensional state space of W x H x θ are respectively (res x , res y , res θ ), and then the device generates discrete three-dimensional pose points in the pre-set area to which the target parking pose point belongs according to the parameters (res x , res y , res θ ), and the first candidate pose state set S is composed of these points. For example, if W is 10 m, H is 5 m, and (res x , res y , res θ ) = (1 m, 1 m, π / 6), then 11 discrete points are generated in the length direction, 6 discrete points are generated in the width direction, and 6 discrete points are generated in the orientation angle. The device will determine a total of 11 * 6 * 6 = 396 three-dimensional pose points, i.e., 396 discrete first candidate pose points.

[0103] Through this embodiment, different length direction step sizes, width direction step sizes, and angle direction step sizes can be set according to actual needs, and different numbers and uniformly distributed first candidate pose points can be selected in the pre-set area to which the target parking pose point belongs.

[0104] In actual applications, there can be multiple pose points near the target parking pose point that have collision-free paths to the target parking pose point. Therefore, in an embodiment, a pose point corresponding to the optimal collision-free path to the target parking pose point can be selected as the collision-free pose point. That is, the collision-free path from the collision-free pose point to the target parking pose point is the collision-free path with the lowest travel cost among the collision-free paths from each pose point in the preset region to the target parking pose point.

[0105] Therefore, in an embodiment, step 201 can specifically include the following steps:

[0106] Step 2011: determining a plurality of first candidate pose points in a preset region to which the target parking pose point belongs, according to a preset pose point selection strategy.

[0107] The description of this step can refer to the description of step 201a.

[0108] Step 2013: for each first candidate pose point, if there is a collision-free path from the first candidate pose point to the target parking pose point, regarding the first candidate pose point as a second candidate pose point.

[0109] The description of this step can refer to the description of step 201b.

[0110] Step 2015: for each second candidate pose point, determining a travel cost of the second candidate pose point traveling to the target parking pose point.

[0111] In an embodiment, when the device determines the travel cost of each second candidate pose point traveling to the target parking pose point, the length of the collision-free path between the second candidate pose point and the target parking pose point can be directly taken as the travel cost, and the longer the length of the collision-free path, the greater the travel cost.

[0112] In an embodiment, when the device obtains one or more travel cost parameters of each second candidate pose point traveling to the target parking pose point, the device then determines the travel cost of the second candidate pose point traveling to the target parking pose point according to the obtained one or more travel cost parameters.

[0113] Step 2017: regarding the second candidate pose point with the minimum travel cost as the collision-free pose point.

[0114] In an embodiment, after the device determines the travel cost of each second candidate pose point traveling to the target parking pose point, the second candidate pose point with the minimum travel cost is regarded as the collision-free pose point by comparison.

[0115] In one embodiment, the driving cost of each second candidate pose point can be the same, and thus the device can determine multiple second candidate pose points with minimum driving cost and select one of the second candidate pose points as the collision-free pose point.

[0116] Through this embodiment, among the multiple last collision-free path segments of the parking path, the optimal one is selected, so that the last collision-free path segment of the parking path planned by the device is faster and safer in driving.

[0117] In one embodiment, for each second candidate pose point, the Voronoi potential value of the second candidate pose point can be calculated as the driving cost of the second candidate pose point.

[0118] The Voronoi potential of a grid point can reflect the distance of the grid point from the surrounding obstacles, and the higher the potential, the closer the surrounding obstacles, which is more dangerous from the perspective of collision avoidance safety. Compared with the general artificial potential field, the Voronoi potential field has an advantage in calculating the potential of the narrow channel area formed by the adjacent obstacles. When two obstacles are close to each other, a narrow channel is formed between the two obstacles. If a general potential field is used for calculation, the potential in the narrow channel will be generally high (because the points in the narrow channel are close to the obstacles), which makes it difficult for the search process that tends to low potential to "cross" the narrow channel. If the Voronoi potential is used to calculate the potential of the narrow channel area, there is a low potential area close to 0 in the narrow channel area (usually at a position equal to the distance of the two obstacles), which facilitates the path search to "cross" the narrow channel.

[0119] Voronoi potential ρ v The calculation formula of (x, y) at the point (x, y) is:

[0120]

[0121] where d o is the nearest distance of the point (x, y) to the obstacle, d v is the nearest distance of the point (x, y) to the generalized Voronoi polygon edge (the generalized Voronoi polygon edge can be directly obtained by the MATLAB function voronoi(x, y)), and α is a positive constant, is the maximum effective distance of the potential field, and when , the potential value ρ v (x, y) = 0. It can also be seen from the calculation formula that when the point (x, y) is located on the generalized Voronoi polygon edge (i.e. d v (x, y) = 0), the potential value ρ v (x, y) = 0. The Voronoi potential ρv (x,y) is in the range [0,1], the Voronoi potential value at the obstacle point is maximum, i.e. v (x,y) = 1. The specific potential field parameters a, Then, the Voronoi potential value can be calculated.

[0122] Let the Voronoi potential value of a certain second candidate pose point r be f1, then

[0123] f1 = Voronoi(r x ,r y )

[0124] where r x ,r y are the position coordinates of the pose point r.

[0125] In an embodiment, for each second candidate pose point, the gradient value of the Voronoi potential along the orientation direction of the second candidate pose point can be calculated as the driving cost value of the second candidate pose point.

[0126] The gradient value of the Voronoi potential along the orientation direction of the second candidate pose point can be obtained by first calculating the Voronoi potential and then performing discrete differential value calculation. The role of reducing the gradient value of the Voronoi potential along the orientation direction of the second candidate pose point is to make the path pointed by the orientation angle of the selected second candidate pose point as far away from the surrounding obstacles as possible. For example, in a corridor, a point is located in the middle of the corridor, along the direction of the corridor, the Voronoi potential gradient value is 0, indicating that walking along this direction is free of obstacles and will be farther and farther away from the obstacles; along the direction perpendicular to the corridor, the Voronoi potential gradient value will become larger and larger, indicating that walking along this direction is with obstacles and will be closer and closer to the obstacles.

[0127] Let the gradient value of the Voronoi potential along the orientation direction of a certain second candidate pose point r be f2, then

[0128]

[0129] where rθ is the orientation angle of the second candidate pose point r.

[0130] In an embodiment, for each second candidate pose point, the integral value of the Voronoi potential on the collision-free path from the second candidate pose point to the target parking pose point can be calculated as the driving cost value of the second candidate pose point.

[0131] The Voronoi potential integral value on the collision-free path from the second candidate pose point to the target parking pose point is the sum of the Voronoi potential of the discrete points on the collision-free path from the second candidate pose point to the target parking pose point. Reducing the Voronoi potential integral value on the collision-free path from the second candidate pose point to the target parking pose point can increase the safety of the collision-free path from the collision-free pose point to the target parking pose point.

[0132] Let the Voronoi potential integral value on the collision-free path from a certain second candidate pose point r to the target parking pose point be f3, then

[0133] f3 =∑ p RS r Voronoi(p x ,p y )

[0134] where RS r is the collision-free path from the second candidate pose point r to the target parking pose point, p is a discrete point on the collision-free path, p x ,p y is the x and y position coordinate value of the p point.

[0135] In an embodiment, for each second candidate pose point, the difference between the orientation angle of the intermediate pose point and the second candidate pose point can be calculated as the driving cost value of the second candidate pose point.

[0136] Reducing the difference between the orientation angle of the intermediate pose point and the second candidate pose point can make the selected collision-free pose point reduce the search difficulty of searching a collision-free path directly to the collision-free pose point from the intermediate pose point. The smaller the difference between the orientation angle of the intermediate pose point and the collision-free pose point, the more likely the intermediate pose point and the collision-free pose point can plan a collision-free path close to a straight line.

[0137] Let the difference between the orientation angle of a certain second candidate pose point r and the intermediate pose point be f4, then

[0138]

[0139] where rθ is the orientation angle of the second candidate pose point, and qθ is the orientation angle of the intermediate pose point q.

[0140] In an embodiment, for each second candidate pose point, the device can obtain at least two driving cost parameters of the second candidate pose point:

[0141] a gradient value of the Voronoi potential along a pose direction of the second candidate pose point, an integral value of the Voronoi potential on a path from the second candidate pose point to a target parking pose point, and a pose direction difference value between an intermediate pose point and the second candidate pose point.

[0142] Then, the device obtains the driving cost value of the candidate pose point according to the obtained driving cost parameters and a preset weighting strategy.

[0143] In one embodiment, the driving cost of the second candidate pose point r is denoted as f, and the driving cost f of the second candidate pose point r can be calculated using the following formula:

[0144] f = w1f1 + w2f2 + w3f3 + w4f4

[0145] where w1-w4 are four non-negative weight constants, f1-f4 are respectively a Voronoi potential value of the second candidate pose point, a gradient value of the Voronoi potential along a pose direction of the second candidate pose point, an integral value of the Voronoi potential on a path from the second candidate pose point to a target parking pose point, and a pose direction difference value between an intermediate pose point and the second candidate pose point, and the values of w1-w4 can be selected as appropriate values according to specific requirements, which are not less than 0.

[0146] Wherein, the greater the f is, the greater the driving cost value of the second candidate pose point in actual planning is.

[0147] In one embodiment, step 2015 can specifically include:

[0148] Step 2015a, for each second candidate pose point, obtaining a driving cost parameter of the second candidate pose point.

[0149] Wherein, the driving cost parameter includes one or more of a Voronoi potential value of the second candidate pose point, a gradient value of the Voronoi potential along a pose direction of the second candidate pose point, an integral value of the Voronoi potential on a path from the second candidate pose point to a target parking pose point, and a pose direction difference value between an intermediate pose point and the second candidate pose point.

[0150] Step 2015b, determining a driving cost value of the second candidate pose point according to the obtained driving cost parameter of the second candidate pose point.

[0151] In one embodiment, the device can directly take one of the following parameters of the second candidate pose point as the driving cost value of the second candidate pose point: the Voronoi potential value of the second candidate pose point, the gradient value of the Voronoi potential along the pose direction of the second candidate pose point, the integral value of the Voronoi potential on the path from the second candidate pose point to the target parking pose point, and the pose direction difference value between the intermediate pose point and the second candidate pose point.

[0152] In one embodiment, the device can select two or more of the following parameters of the second candidate pose point: the Voronoi potential value of the second candidate pose point, the gradient value of the Voronoi potential along the pose direction of the second candidate pose point, the integral value of the Voronoi potential on the path from the second candidate pose point to the target parking pose point, and the pose direction difference value between the intermediate pose point and the second candidate pose point, and perform weighted calculation according to actual needs to obtain the driving cost value of the second candidate pose point.

[0153] As shown in FIG. 1, it is a flowchart of the parking path planning method according to one specific embodiment of the present application. Figure 4

[0154] 1. First, determine the current pose point and the target pose point of the vehicle driving in the path planning, and initialize the open set as the current pose point and the closed set as empty.

[0155] 2. In the second candidate pose point set, select the second candidate pose point with the minimum current driving cost value as the collision-free pose point.

[0156] 3. Determine whether the open set is empty. If it is empty, it means that all the pose points of the current map have been searched and no collision-free parking path has been searched, so the path search fails. If it is not empty, continue to perform step 4.

[0157] 4. In the open set, take the pose point with the minimum total cost f as the intermediate pose point.

[0158] 5. Determine whether there is a collision-free path between the intermediate pose point and the collision-free pose point. If there is, the path planning is successful, and the final parking path is produced according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point. If there is not, continue to perform step 6.

[0159] 6. Determine and select a nearby pose point of the intermediate pose point according to the kinematic formula.

[0160] ​7. Determine whether there is a collision-free path from the intermediate pose point to the neighboring pose point. If there is no collision-free path, return to step 6 to select another neighboring pose point. If there is a collision-free path, continue to step 8.

[0161] 8. Determine whether the neighboring pose point is in the closed set. If the neighboring pose point is in the closed set, it means that the neighboring pose point has been traversed and there is no need to recompute the cost associated with the neighboring pose point. Return to step 6 to select another neighboring pose point. If the neighboring pose point is not in the closed set, continue to step 9.

[0162] 9. Calculate the cost g of moving from the current pose point to the neighboring pose point based on the cost of moving from the current pose point to the intermediate pose point and the cost of moving from the intermediate pose point to the neighboring pose point.

[0163] 10. Determine whether the neighboring pose point is in the open set. If yes, continue to step 11. If no, continue to step 13.

[0164] 11. Determine whether the cost g of the neighboring pose point is less than the cost g of the neighboring pose point stored in the open set. If the cost g of the neighboring pose point is not less than the cost g of the neighboring pose point stored in the open set, return to step 2. If the cost g of the neighboring pose point is less than the cost g of the neighboring pose point stored in the open set, continue to step 12.

[0165] 12. Update the cost g of the neighboring pose point, the parent node of the neighboring pose point, and the total cost f of the neighboring pose point in the open set. Return to step 2.

[0166] 13. Calculate the heuristic cost h of the neighboring pose point. Calculate the total cost f of the neighboring pose point based on the cost g and the heuristic cost h of the neighboring pose point. Record the parent node of the neighboring pose point as the intermediate pose point. Add the neighboring pose point to the open set. Return to step 2.

[0167] The above is the detailed description of the parking path planning method of the present application. It should be understood that although each step in the flowchart involved in each embodiment described above is displayed in sequence according to the arrow, these steps are not necessarily executed in the order indicated by the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in each embodiment described above can include multiple steps or stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but can be alternately executed with at least part of other steps or stages or steps or stages in other steps.

[0168] Based on the same inventive concept, the embodiments of the present application also provide a parking path planning device for implementing the above-mentioned parking path planning method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more parking path planning device embodiments provided below can refer to the limitations of the parking path planning method described above, and will not be repeated here.

[0169] In one embodiment, as shown in Figure 5 a parking path planning device is provided, comprising: an acquisition module 501, a determination module 503 and a generation module 505, wherein:

[0170] The acquisition module 501 is configured to acquire a collision-free pose point in a preset area to which a target parking pose point belongs; the collision-free pose point is a pose point having a collision-free path with the target parking pose point;

[0171] The determination module 503 is configured to determine whether there is a collision-free path from an intermediate pose point to the collision-free pose point; the intermediate pose point is a pose point between a current pose point and the target parking pose point determined according to a preset path search strategy;

[0172] The generation module 505 is configured to, if there is, generate a parking path according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point.

[0173] In one embodiment, the above-mentioned parking path planning device further comprises:

[0174] The updating module 507 (not shown in the figure) is configured to re-determine the intermediate pose point according to the path searching strategy if there is no collision-free path from the intermediate pose point to the collision-free pose point, and call the determining module 503.

[0175] In one embodiment, the obtaining module 501 described above specifically comprises a selecting sub-module 5011 (not shown in the figure) and a screening sub-module 5013 (not shown in the figure), wherein,

[0176] The selecting sub-module 5011 is configured to determine a plurality of first candidate pose points in a preset area to which the target parking pose point belongs, according to a preset pose point selection strategy.

[0177] The screening sub-module 5013 is configured to, for each first candidate pose point, take the first candidate pose point as a collision-free pose point if there is a collision-free path from the first candidate pose point to the target parking pose point.

[0178] In one embodiment, the selecting sub-module 5011 described above specifically comprises a position point determining sub-unit 5011a (not shown in the figure) and a pose point determining sub-unit 5011b (not shown in the figure), wherein,

[0179] The position point determining sub-unit 5011a is configured to determine a plurality of position points in the preset area based on a preset position selection strategy.

[0180] The pose point determining sub-unit 5011b is configured to determine a pose point corresponding to each position point as a first candidate pose point based on a preset pose angle.

[0181] In one embodiment, the obtaining module 501 described above specifically comprises a first candidate pose point determining sub-module 501a (not shown in the figure), a second candidate pose point determining sub-module 501b (not shown in the figure), a driving cost value determining sub-module 501c (not shown in the figure), a collision-free pose point determining sub-module 501d (not shown in the figure), wherein,

[0182] The first candidate pose point determining sub-module 501a is configured to determine a plurality of first candidate pose points in a preset area to which the target parking pose point belongs, according to a preset pose point selection strategy.

[0183] The second candidate pose point determining sub-module 501b is configured to, for each first candidate pose point, take the first candidate pose point as a second candidate pose point if there is a collision-free path from the first candidate pose point to the target parking pose point.

[0184] The driving cost value determination submodule 501c is configured to determine, for each second candidate pose point, a driving cost value of the second candidate pose point to the target parking pose point.

[0185] The collision-free pose point determination submodule 501d is configured to determine the second candidate pose point with the minimum driving cost value as the collision-free pose point.

[0186] In an embodiment, the driving cost value determination submodule 501c is specifically configured to:

[0187] For each second candidate pose point, obtain a driving cost parameter of the second candidate pose point, wherein the driving cost parameter comprises one or more of a Voronoi potential value of the second candidate pose point, a gradient value of the Voronoi potential along a pose direction of the second candidate pose point, a Voronoi potential integral value on a path from the second candidate pose point to the target parking pose point, and a pose direction difference value between the intermediate pose point and the second candidate pose point.

[0188] According to the obtained driving cost parameter of the second candidate pose point, determine the driving cost value of the second candidate pose point.

[0189] Each of the above modules in the parking path planning device can be realized by software, hardware, or a combination thereof, in whole or in part. Each of the above modules can be embedded in or independent of a processor in a computer device in hardware form, or can be stored in a memory in the computer device in software form, so as to be called and executed by a processor to perform the operations corresponding to each of the above modules.

[0190] In an embodiment, a computer device is provided, which can be a terminal, and an internal structure diagram of the computer device can be as shown in FIG. 8. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved through WIFI, mobile cellular network, NFC (near field communication), or other technologies. The computer program is executed by the processor to implement a parking path planning method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or can be a key, trackball, or touchpad arranged on the shell of the computer device, or can be an external keyboard, touchpad, or mouse, etc. Figure 6 In an embodiment, a computer device is provided, which can be a terminal, and an internal structure diagram of the computer device can be as shown in FIG. 8. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved through WIFI, mobile cellular network, NFC (near field communication), or other technologies. The computer program is executed by the processor to implement a parking path planning method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or can be a key, trackball, or touchpad arranged on the shell of the computer device, or can be an external keyboard, touchpad, or mouse, etc.

[0191] Those skilled in the art can understand that Figure 6 Those skilled in the art can understand that

[0192] In an embodiment, a computer device is also provided, including a memory and a processor, the memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0193] In an embodiment, a computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the steps in the above method embodiments.

[0194] In an embodiment, a computer program product is provided, which includes a computer program, and the computer program is executed by a processor to implement the steps in the above method embodiments.

[0195] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric memory (Ferroelectric Random Access Memory, FRAM), phase change memory (Phase Change Memory, PCM), graphene memory, etc. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0196] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0197] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A parking path planning method characterized by comprising: The method comprises: acquiring a collision-free pose point in a preset region to which a target parking pose point belongs; the collision-free pose point is a pose point having a collision-free path with the target parking pose point; determining whether a collision-free path exists between an intermediate pose point and the collision-free pose point; the intermediate pose point is a pose point between a current pose point and the target parking pose point, which is determined according to a preset path search strategy; if a collision-free path exists, generating a parking path according to the collision-free path from the current pose point to the intermediate pose point, the collision-free path from the intermediate pose point to the collision-free pose point, and the collision-free path from the collision-free pose point to the target parking pose point; the acquiring of the collision-free pose point in the preset region to which the target parking pose point belongs comprises: determining a plurality of first candidate pose points in the preset region to which the target parking pose point belongs, according to a preset pose point selection strategy; for each first candidate pose point, if a collision-free path exists between the first candidate pose point and the target parking pose point, taking the first candidate pose point as a second candidate pose point; for each second candidate pose point, determining a driving cost value of the second candidate pose point driving to the target parking pose point; taking the second candidate pose point having the minimum driving cost value as the collision-free pose point; the determining of the driving cost value of the second candidate pose point driving to the target parking pose point comprises: for each second candidate pose point, acquiring a driving cost parameter of the second candidate pose point; wherein the driving cost parameter comprises one or more of a Voronoi potential value of the second candidate pose point, a gradient value of the Voronoi potential along a pose direction of the second candidate pose point, a Voronoi potential integral value on a path from the second candidate pose point to the target parking pose point, and a pose direction difference value between the intermediate pose point and the second candidate pose point; determining the driving cost value of the second candidate pose point according to the acquired driving cost parameter of the second candidate pose point.

2. The method of claim 1, wherein, The method further comprises: if a collision-free path does not exist between the intermediate pose point and the collision-free pose point, re-determining the intermediate pose point according to the path search strategy, and performing the step of determining whether a collision-free path exists between the intermediate pose point and the collision-free pose point.

3. The method of claim 1, wherein, the determining of the driving cost value of the second candidate pose point according to the acquired driving cost parameter of the second candidate pose point comprises: selecting two or more driving cost parameters from the Voronoi potential value of the second candidate pose point, the gradient value of the Voronoi potential along the pose direction of the second candidate pose point, the Voronoi potential integral value on the path from the second candidate pose point to the target parking pose point, and the pose direction difference value between the intermediate pose point and the second candidate pose point, and performing weighted calculation to obtain the driving cost value of the second candidate pose point.

4. The method of claim 1, wherein, The Voronoi potential of the grid point can reflect the distance between the grid point and the surrounding obstacles, and the higher the Voronoi potential is, the closer the surrounding obstacles are.

5. The method of claim 1, wherein, The integral value of the Voronoi potential on the collision-free path from the second candidate pose point to the target parking pose point is the sum of the Voronoi potentials of the discrete points on the collision-free path from the second candidate pose point to the target parking pose point, and reducing the integral value of the Voronoi potential on the collision-free path from the second candidate pose point to the target parking pose point can increase the safety of the selected collision-free pose point on the collision-free path from the collision-free pose point to the target parking pose point.

6. The method of claim 1, wherein, Reducing the difference between the orientation angles of the intermediate pose point and the second candidate pose point can make the selected collision-free pose point reduce the search difficulty of searching a collision-free path directly reaching the collision-free pose point from the intermediate pose point.

7. A parking route planning apparatus characterized by comprising: The device comprises: an acquisition module configured to acquire a collision-free pose point in a preset region to which a target parking pose point belongs; the collision-free pose point is a pose point having a collision-free path with the target parking pose point; a determination module configured to determine whether there is a collision-free path from an intermediate pose point to the collision-free pose point; the intermediate pose point is a pose point between a current pose point and the target parking pose point determined according to a preset path search strategy; a generation module configured to generate a parking path according to a collision-free path from the current pose point to the intermediate pose point, a collision-free path from the intermediate pose point to the collision-free pose point, and a collision-free path from the collision-free pose point to the target parking pose point if there is the collision-free path from the intermediate pose point to the collision-free pose point; the acquisition module comprises a first candidate pose point determination submodule configured to determine a plurality of first candidate pose points in a preset region to which a target parking pose point belongs according to a preset pose point selection strategy; a second candidate pose point determination submodule configured to, for each first candidate pose point, take the first candidate pose point as a second candidate pose point if there is a collision-free path from the first candidate pose point to the target parking pose point; a driving cost value determination submodule configured to determine a driving cost value of driving of each second candidate pose point to the target parking pose point; a collision-free pose point determination submodule configured to take a second candidate pose point having a minimum driving cost value as a collision-free pose point; and a path generation submodule configured to generate a parking path according to a collision-free path from the current pose point to the intermediate pose point, a collision-free path from the intermediate pose point to the collision-free pose point, and a collision-free path from the collision-free pose point to the target parking pose point if there is the collision-free path from the intermediate pose point to the collision-free pose point. The driving cost value determination submodule is specifically configured to obtain, for each second candidate pose point, a driving cost parameter of the second candidate pose point, wherein the driving cost parameter comprises one or more of a Voronoi potential value of the second candidate pose point, a gradient value of the Voronoi potential along a pose direction of the second candidate pose point, a Voronoi potential integral value on a path from the second candidate pose point to the target parking pose point, and a pose direction difference value between the intermediate pose point and the second candidate pose point; and determine a driving cost value of the second candidate pose point according to the obtained driving cost parameter of the second candidate pose point.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor, when executing the computer program, implements the steps of the method of any one of claims 1 to 6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by the processor, implements the steps of the method of any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program, when executed by the processor, implements the steps of the method of any one of claims 1 to 6. The computer program, when executed by the processor, implements the steps of the method of any one of claims 1 to 6.

Images