Self-adaptive path planning method for autonomous vehicle

Abstract

The invention relates to the technical field of automatic driving, and particularly provides a self-adaptive path planning method for an automatic driving vehicle, and the method comprises the steps: selecting an exhibition point based on a current position point of a target vehicle and a pre-planned initial global path; planning a smooth path according to the current position point and the exhibition point of the target vehicle; the initial global path is updated based on the smooth path, a target global path is obtained, and the target global path comprises the smooth path from the current position point to the exhibition point and the initial global path after the exhibition point; performing collision detection on the target global path; and under the condition that the collision detection is passed, navigating the target vehicle based on the target global path. The problem that a path planning system in the related technology cannot give consideration to high vehicle driving efficiency, high real-time performance and short calculation time is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous driving, and particularly to an adaptive path planning method for an autonomous vehicle. Background Art

[0002] Path planning is one of the key technologies for autonomous vehicles to drive safely and efficiently. It needs to navigate the vehicle in a complex traffic environment, and be able to handle various static and dynamic obstacles, especially in urban or industrial areas with a large number of large-scale static obstacles. In the field of autonomous driving technology, path planning systems are mainly divided into three categories: reactive, sampling-based, and refinement-based.

[0003] The reactive planning system, also known as the obstacle avoidance system, can quickly respond to sudden changes in the environment, such as suddenly appearing obstacles, but often can only find a local optimal solution and cannot find the global optimal path from the starting point to the end point, resulting in low vehicle driving efficiency.

[0004] The sampling-based planning system, such as using the RRT (Rapidly-exploring Random Tree) algorithm, does not rely on the initial path, but searches for random sampling points in the search space and constructs a tree structure to find the path, and can find the global optimal solution, but requires a large amount of computing resources, especially in high-dimensional spaces, and is not efficient enough in real-time applications.

[0005] The refinement-based planning system finds the optimal solution by iteratively optimizing the global path, and can find a path with a lower cost, but requires a long computing time and cannot respond to dynamic changes in the environment in a timely manner.

[0006] Regarding the problem that the path planning systems in the related technologies cannot balance high vehicle driving efficiency, high real-time performance, and short computing time, no effective solution has been proposed yet. Summary of the Invention

[0007] An adaptive path planning method for an autonomous vehicle provided by an embodiment of the present invention can at least solve the problem that the path planning systems in the related technologies cannot balance high vehicle driving efficiency, high real-time performance, and short computing time.

[0008] An adaptive path planning method for an autonomous vehicle provided by an embodiment of the present invention includes: selecting a prospective point based on the current position point of the target vehicle and the initially planned initial global path; planning a smooth path according to the current position point and the prospective point; updating the initial global path based on the smooth path to obtain a target global path, where the target global path includes the smooth path from the current position point to the prospective point and the initial global path after the prospective point; performing a collision detection on the target global path; and navigating the target vehicle based on the target global path when the collision detection passes.

[0009] The adaptive path planning method for an autonomous vehicle provided by an embodiment of the present invention, after navigating the target vehicle based on the target global path, further includes: in the next planning cycle, using the target global path as the initial global path and re-planning to obtain the corresponding target global path; wherein, the above-mentioned planning cycle is a planning cycle whose duration and frequency can both be adjusted.

[0010] The adaptive path planning method for an autonomous vehicle provided by an embodiment of the present invention further includes: in the case where the collision detection fails, determining the collision point of the target global path; calculating the collision rate of the collision point, and determining the correction direction according to the collision rate; determining the correction angle based on the correction direction; and correcting the target global path according to the correction angle.

[0011] The adaptive path planning method for an autonomous vehicle provided by an embodiment of the present invention, calculating the collision rate of the collision point and determining the correction direction according to the collision rate, includes: respectively calculating the left collision overlapping area and the right collision overlapping area between the target vehicle and the collision point; calculating the left collision rate based on the left collision overlapping area and vehicle parameters, and calculating the right collision rate based on the right collision overlapping area and vehicle parameters, wherein the vehicle parameters include the vehicle length and the vehicle width; correcting to the right in the case where the left collision rate is greater than the right collision rate; and correcting to the left in the case where the right collision rate is greater than the left collision rate.

[0012] The adaptive path planning method for an autonomous vehicle provided by an embodiment of the present invention, determining the correction angle based on the correction direction, includes: determining the correction angle based on the angle between the first direction and the orientation of the target vehicle, and the correction direction; wherein, the first direction is the direction from the current position point of the target vehicle to the collision point.

[0013] The adaptive path planning method for an autonomous vehicle provided by an embodiment of the present invention, correcting the target global path according to the correction angle, includes: correcting the coordinates of the collision point based on the correction angle to obtain the coordinates of the corrected point; and correcting the target global path based on the corrected point and the points other than the collision point in the target global path to obtain the corrected global path.

[0014] After the adaptive path planning method for an autonomous vehicle provided by an embodiment of the present invention corrects the target global path to obtain the corrected global path, it further includes: determining the smoothed path corresponding to the corrected global path and the target global path, and performing collision detection; in the case where the collision detection fails, continuing to correct until the final target global path passes the collision detection.

[0015] The adaptive path planning method for an autonomous driving vehicle provided by an embodiment of the present invention creates a smooth path based on an initial global path to select a prospective point and according to the current position point and the prospective point of the target vehicle, and includes: determining the radius of an arc based on the distance between the prospective point and the current position point, and the included angle between the second direction and the orientation of the target vehicle, where the second direction is the direction from the current position point to the prospective point; using the arc as the smooth path.

[0016] The adaptive path planning method for an autonomous driving vehicle provided by an embodiment of the present invention creates a navigation for the target vehicle based on a target global path, and includes: calculating a steering angle based on the radius of the smooth path; navigating the target vehicle based on the steering angle within a preset time, where the preset time is the time for the target vehicle to track the prospective point.

[0017] Before selecting a prospective point by a sampling-based planning method through the current position point of the target vehicle and a pre-planned initial global path, the adaptive path planning method for an autonomous driving vehicle provided by an embodiment of the present invention creates also includes: generating a grid map based on the initial position information, target position information, and obstacle information of the target vehicle; planning an initial global path in the plane rectangular coordinate system of the grid map through a path planning algorithm.

[0018] An embodiment of the present invention also provides an electronic device, including: a processor, and a memory storing a program, characterized in that the program includes instructions that, when executed by the processor, cause the processor to execute the above-mentioned adaptive path planning method for an autonomous driving vehicle.

[0019] The adaptive path planning method for an autonomous driving vehicle provided by an embodiment of the present invention creates a prospective point by using a sampling-based planning method through an initial global path and a current position point, plans a precise smooth path through a refinement-based planning method within the local area between the prospective point and the current position point, updates the initial global path through the smooth path, determines a target global path, and navigates the target vehicle. While global planning can be achieved in each planning cycle, precise planning of the current position can be performed, which can not only ensure the driving efficiency of the vehicle, but also ensure the efficiency and speed of path planning, and reduce the data processing volume of refinement planning. Combining the setting of the planning cycle, the above-mentioned planning can be performed in each planning cycle, thereby realizing dynamic planning of the target vehicle. Thus, it solves the problem of low driving efficiency in the prior art when simply using the sampling-based planning method, and the problems of long calculation time and low planning efficiency when simply using the refinement-based planning method.

[0020] After determining the target global path, collision detection can also be performed. When the collision detection passes, the target vehicle is navigated based on the target global path that has passed the collision detection, which can improve the vehicle driving efficiency and respond in a timely manner to the dynamic changes in the environment. In summary, the above method of the present invention solves the problem in the related art that the path planning system cannot take into account high vehicle driving efficiency, high real-time performance, and short calculation time. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other embodiments based on these drawings without creative efforts.

[0022] Figure 1 It is a flowchart of the steps of an adaptive path planning method for an autonomous vehicle in an embodiment of the present invention.

[0023] Figure 2 It is a schematic diagram of planning a smooth path according to the current position point and the look-ahead point in an embodiment of the present invention.

[0024] Figure 3 It is a schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] The embodiments of the present invention will be described in more detail below with reference to the drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Instead, these embodiments are provided to more thoroughly and completely understand the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not used to limit the protection scope of the present invention.

[0026] In order to solve the above technical problems, an embodiment of the present invention provides an adaptive path planning method for an autonomous vehicle.

[0027] Please refer to Figure 1 As shown, the above method includes:

[0028] Step S101, based on the current position point of the target vehicle and the initially planned initial global path, select a look-ahead point through a sampling-based planning method;

[0029] Step S102, plan a smooth path according to the current position point and the look-ahead point of the target vehicle through a refinement-based planning method;

[0030] Step S103: Update the initial global path based on the smooth path to obtain the target global path, where the target global path includes the smooth path from the current position point to the look-ahead point and the initial global path after the look-ahead point.

[0031] Step S104: Perform a collision detection on the target global path.

[0032] Step S105: When the collision detection passes, navigate the target vehicle based on the target global path.

[0033] It can be understood that the target vehicle is an autonomous vehicle to be navigated, and a navigation system is installed on the vehicle. The navigation system applies the above method provided by the embodiment of the present invention.

[0034] In a plane rectangular coordinate system, the initial global path is composed of a series of path points, and the look-ahead point is a path point used to plan the smooth path. A smooth path can be planned according to the current position point of the target vehicle and any look-ahead point.

[0035] The above look-ahead point can be a position point within a preset distance range from the current position point. When selecting the look-ahead point corresponding to the current position point on the initial global path, the look-ahead point can be selected by random sampling on the initial global path based on the set distance range from the current position point.

[0036] The target global path consists of two parts. One part is the smooth path planned by the refined navigation method. The starting point of this smooth path is the current position point of the target vehicle, and the ending point is the look-ahead point corresponding to this smooth path. The other part is the part of the initial global path starting from the above look-ahead point until the end point, and the end point refers to the point corresponding to the above target position information in the plane rectangular coordinate system.

[0037] During the driving process of the target vehicle, the current position point of the target vehicle, as well as the look-ahead point and the smooth path corresponding to the current position point, are dynamically updated in real time through the set planning period. The selection of the look-ahead point can be through a sampling-based planning method, and random sampling is performed in the state space determined by the current position point.

[0038] The above state space can be a space composed of states such as the position and attitude of the target vehicle. The above sampling-based planning method can be the Rapidly - Exploring Random Tree (RRT).

[0039] The range of the smooth path is small, and the amount of data for refined navigation is small, so an accurate navigation path can be planned relatively quickly. In this way, the path can be dynamically planned following the target vehicle and navigation can be performed, enabling the target vehicle to always travel along the accurate navigation path during driving.

[0040] The planning method of the smooth path in this embodiment will be specifically described later. The above-mentioned planning method of the initial global path can be obtained by any planning method such as the above-mentioned reactive planning system, sampling planning system, refined planning system, etc. The initial global path only needs to be generated once. Therefore, before the target vehicle travels, the refined planning system can be used for refined planning to obtain an accurate initial global path, providing an accurate basis for subsequent local smooth path planning. For example, the initial global path is planned by the A* algorithm.

[0041] Compared with the refined navigation method in the related art, which requires accurate navigation of the global path, not only is the amount of data processed large, but also the efficiency of path planning and navigation is low. Moreover, the method of pre-planning the global path is difficult to effectively avoid dynamic obstacles in the global path.

[0042] In this embodiment, by dynamically planning the refined path within a predetermined range of the current position of the target vehicle, not only is the amount of data processed small, but also the efficiency of path planning and navigation is high, and dynamic obstacles in the path can be effectively avoided, greatly improving the accuracy of navigation.

[0043] Based on the initial global path and the current position point, the sampling planning method is used to select the prospective point, and the refined planning method is used to plan within the local area between the prospective point and the current position point to obtain an accurate smooth path. The initial global path is updated through the smooth path to determine the target global path for navigating the target vehicle.

[0044] While global planning can be achieved in each planning cycle, accurate planning of the current position can not only ensure the driving efficiency of the vehicle, but also ensure the efficiency and speed of path planning, reducing the data processing volume of refined planning.

[0045] It can also be combined with the setting of the planning cycle, and the above-mentioned planning can be performed in each planning cycle, thereby realizing the dynamic planning of the target vehicle. Thus, it solves the problem of relatively low driving efficiency in the prior art when simply using the sampling planning method, and the problems of long calculation time and low planning efficiency when simply using the refined planning method.

[0046] It should be noted that when setting multiple planning cycles, the above steps S101 to S105 can be performed in each planning cycle.

[0047] The above-mentioned collision detection is mainly divided into three categories: collision detection based on sensor data, collision detection based on map information, and collision detection based on kinematic simulation.

[0048] Exemplarily, a method of processing lidar point cloud is used for collision detection based on sensor data. The lidar can obtain high-precision three-dimensional point cloud data. The point cloud data is clustered, and points with similar distances are grouped into one group. Each group of points represents a possible obstacle. The bounding box of the obstacle is calculated, and whether a collision will occur is judged by comparing the planned path of the target vehicle with the position relationship of the bounding box.

[0049] For example, if the planned path of the target vehicle intersects with the bounding box of the obstacle, there is a risk of collision.

[0050] In other words, whether a collision will occur is judged by comparing the position relationship between the target global path and the obstacle in the plane rectangular coordinate system. When the target global path does not intersect with the obstacle, the collision detection passes.

[0051] Collision detection can also be performed after determining the target global path. When the collision detection passes, the target vehicle is navigated based on the target global path that has passed the collision detection, which can improve the driving efficiency of the target vehicle and respond to the dynamic changes of the environment in a timely manner. In summary, the above method of the present invention solves the problem in the related art that the path planning system cannot take into account both high driving efficiency, high real-time performance, and short calculation time of the vehicle.

[0052] When the above-mentioned collision detection fails, a reactive planning method can be used to update and navigate the obstacle avoidance path, so as to effectively avoid dynamic obstacles and greatly improve the rationality of path planning.

[0053] In combination with the planning cycle, multiple planning cycles can dynamically adjust the target global path. When a dynamic obstacle generates an obstacle with the global path, the target global path is avoided from the obstacle. When the obstacle moves dynamically and does not generate an obstacle, the target global path is planned back to the original position, achieving the purpose of dynamically avoiding dynamic obstacles and realizing the most reasonable current plan at the current position.

[0054] Preferably, after navigating the target vehicle based on the target global path, the above method further includes:

[0055] In the next planning cycle, the target global path is used as the initial global path for re-planning to obtain the corresponding target global path; wherein, the above-mentioned planning cycle is a planning cycle with adjustable duration and frequency.

[0056] It can be understood that the number of planning cycles corresponds to the number of path planning times. The accuracy of the target global path can be improved through multiple path planning. The target global path obtained in the current planning cycle is used as the initial global path for the next planning cycle to achieve adaptive iteration and improve the smoothness of the finally obtained target global path.

[0057] It can be understood that after the above steps S101 - S105 are performed in the current planning cycle, in the next planning cycle, the target global path of the current planning cycle can be used as the initial global path for the next planning cycle, and steps S101 - S105 are re - executed for its corresponding initial global path to achieve real - time dynamic planning of the target vehicle.

[0058] In the scenario of real - time dynamic planning, the above method can achieve refined planning of the current position in each planning cycle, so that the target vehicle can always navigate along the accurate and smooth path of the refined planning during driving, but it does not require refined planning of the entire global path each time, thereby improving the accuracy rate, planning efficiency, and real - time performance of path planning.

[0059] Those skilled in the art can determine the duration and frequency of the planning cycle according to empirical values and reference values of the system or device.

[0060] Preferably, the duration of the above - mentioned planning cycle can be 1s - 3s. Adjacent planning cycles are connected end - to - end without setting an interval between cycles to achieve real - time dynamic planning.

[0061] Preferably, in step S101, before selecting the look - ahead point through the sampling - type planning method based on the current position point of the target vehicle and the pre - planned initial global path, it is necessary to plan the initial global path in the plane rectangular coordinate system based on the initial position information, target position information, and obstacle information of the target vehicle. Specifically:

[0062] Feature extraction is performed on the obstacle information to obtain feature information.

[0063] Based on the initial position information, target position information, and feature information of the target vehicle, the initial global path is planned in the plane rectangular coordinate system.

[0064] Among them, in the case where the driving environment of the target vehicle is relatively complex, such as a narrow street with a large number of non - motor vehicles and pedestrians, the obstacle information includes semantic information of the street - crossing scenario.

[0065] In other words, in a relatively good driving environment, obstacle information can be simplified to necessary information such as road infrastructure to improve the planning efficiency. However, in a relatively poor driving environment, the accuracy of the planned target global path can be improved by increasing obstacle information to enhance driving safety.

[0066] It can be understood that the initial position information and target position information of the vehicle are determined by the navigation system carried by the target vehicle, and obstacle information is obtained through sensors or vehicle networking technology.

[0067] Sensors include but are not limited to lidar, millimeter-wave radar, cameras, and ultrasonic sensors.

[0068] Vehicle networking technology refers to the communication between the target vehicle and other vehicles or road infrastructure around it. Other vehicles can send the obstacle information detected by themselves to the target vehicle. Road infrastructure, such as intelligent traffic lights and roadside sensors, can send the monitored relevant information, such as the appearance of a malfunctioning vehicle at an intersection, to the target vehicle.

[0069] Obviously, the combined use of sensors and vehicle networking technology can obtain more accurate obstacle information.

[0070] It can be understood that the current position information of the vehicle can be determined according to the navigation system carried by the target vehicle, and the current position point is determined in the plane rectangular coordinate system based on the current position information.

[0071] Preferably, the above-mentioned initial global path is planned in the plane rectangular coordinate system based on the initial position information, target position information, and obstacle information of the target vehicle, including:

[0072] Generate a grid map based on the initial position information, target position information, and obstacle information of the target vehicle.

[0073] Construct a plane rectangular coordinate system based on the grid map.

[0074] Plan the initial global path in the plane rectangular coordinate system through a path planning algorithm.

[0075] It can be understood that the grid map can divide the environmental space into grids with a series of self-defined rules. Each grid can store corresponding various attribute information. For example, set a specific value or symbol to specially mark the grid corresponding to the initial position information of the target vehicle, and perform similar processing on the grid corresponding to the target position information, so that the subsequent path planning algorithm can identify the starting point and the ending point.

[0076] Select the vertex of a certain grid as the origin of the plane rectangular coordinate system, and establish the plane rectangular coordinate system accordingly. For example, the range is set such that the x-axis ranges from 0 to 100 meters and the y-axis ranges from 0 to 100 meters, and the grid resolution is selected as 1 meter × 1 meter. There are 100 × 100 grids that can be represented by coordinates.

[0077] Exemplarily, an initial global path GP is planned in the plane rectangular coordinate system through the A* algorithm. Figure 2 The local path of the initial global path GP is shown.

[0078] Among them, the center or vertex of each of the above 100 × 100 grids is regarded as a node, and these nodes constitute the search space of the A* algorithm.

[0079] Preferably, in step S101, based on the current position point of the target vehicle and the pre-planned initial global path, prospect points are selected through a sampling-based planning method, including:

[0080] Sample the position points on the initial global path with a preset probability to obtain prospect points.

[0081] Alternatively, estimate the cost between the current position point and each position point on the initial global path, sort the position points on the initial global path in ascending order of cost, and take the position points ranked in the front as prospect points. For example, take the position points ranked in the top 50% as prospect points.

[0082] Among them, the cost can refer to the distance between the current position point and each position point on the initial global path, such as the Euclidean distance or the Manhattan distance; the cost can also refer to the time cost, the energy consumption cost, and the safety risk cost. Those skilled in the art can select the specific type of cost according to actual application requirements.

[0083] It can be understood that sampling prospect points based on a preset probability and selecting prospect points based on cost each have their advantages and disadvantages. The former has a relatively small computational amount but relatively low real-time performance, while the latter has a relatively large computational amount but relatively high real-time performance.

[0084] Preferably, in step S102, a smooth path is planned through a refinement-based planning method according to the current position point of the target vehicle and the prospect points, including:

[0085] Based on the distance between the prospect point and the current position point, and the included angle between the second direction and the orientation of the target vehicle, determine the radius of the arc, where the second direction is the direction from the current position point to the prospect point.

[0086] Take the above arc as the smooth path.

[0087] The above radius R is determined by the following formula:

[0088]

[0089] Please refer to Figure 2 as shown, which represents the distance between the prospective point Q and the current position point P, and represents the angle between the second direction and the orientation of the target vehicle. The second direction is the direction from point P to point Q in Figure 2 and the arc between the current position point P and the prospective point Q is the smooth path SP.

[0090] In the above manner, the smooth path between the current position point and the prospective point can be determined. Not only is the finally generated path smooth without sharp turns, but the calculation is also convenient and fast, and it is also convenient to determine the function of the steering angle to navigate the target vehicle.

[0091] Preferably, in step S105, navigating the target vehicle based on the target global path includes:

[0092] Calculating the steering angle based on the radius of the smooth path.

[0093] Navigating the target vehicle based on the steering angle within a preset time, where the preset time is the time for the target vehicle to track the prospective point.

[0094] Steering angle function is as follows:

[0095]

[0096] In the formula, represents the distance between the current position point P and the position point at time t.

[0097] According to the above steering angle function, multiple times after the current time corresponding to the current position of the target vehicle are determined, and based on the time difference between the corresponding time and the current time, the steering angles corresponding to the subsequent times are calculated.

[0098] Thereby, a data sequence of multiple times and the corresponding steering angles after the current time is generated, and this data sequence is sent to the target vehicle to control the target vehicle to turn at the corresponding steering angle at the corresponding time, thereby realizing the navigation of the target vehicle.

[0099] Exemplarily, the preset time is set to one second. Within this one second, multiple times can be generated, and the target vehicle is navigated based on the steering angles corresponding to each time, so that the target vehicle travels along the smooth path SP.

[0100] Preferably, in step S105, the above method further includes:

[0101] Step S1051, in the case where the collision detection fails, determining the collision point of the target global path.

[0102] Step S1052, calculate the collision rate of the collision point, and determine the correction direction according to the collision rate.

[0103] Step S1053, determine the correction angle based on the correction direction.

[0104] Step S1054, correct the target global path according to the correction angle.

[0105] Among them, the collision point is a point on the current target global path, and the attribute information of this point on the grid map conflicts with the obstacle information, indicating that the target vehicle will collide with the obstacle at the collision point when driving according to the current target global path.

[0106] It can be understood that the obstacle information in the actual scenario is dynamically changing. Therefore, it is necessary to perform collision detection on the target global path and make corrections when the collision detection fails, which can improve the navigation safety.

[0107] Preferably, in step S1052, calculating the collision rate of the collision point and determining the correction direction according to the collision rate includes:

[0108] Calculate the left collision overlap area and the right collision overlap area between the target vehicle and the collision point respectively.

[0109] Calculate the left collision rate based on the left collision overlap area and vehicle parameters, and calculate the right collision rate based on the right collision overlap area and vehicle parameters, where the vehicle parameters include the vehicle length and vehicle width.

[0110] When the left collision rate is greater than the right collision rate, correct to the right.

[0111] When the right collision rate is greater than the left collision rate, correct to the left.

[0112] Exemplarily, calculate the left collision overlap area and the right collision overlap area in the above grid map based on the A* algorithm. Among them, the target vehicle is represented by a rectangle in the above grid map.

[0113] Calculate the left collision rate respectively through the following formula and the right collision rate :

[0114]

[0115]

[0116] In the formula, represents the left collision overlap area, represents the right collision overlap area, represents the vehicle length of the target vehicle, Represents the vehicle width of the target vehicle.

[0117] Preferably, in step S1053, determining a correction angle based on the correction direction includes:

[0118] Determining a correction angle based on the angle between the first direction and the orientation of the target vehicle, and the correction direction.

[0119] Wherein, the first direction is the direction from the current position point of the target vehicle to the collision point.

[0120] When the correction direction is a right correction, the correction angle is determined by the following formula :

[0121] ;

[0122] In the formula, Represents the angle between the first direction and the orientation of the target vehicle.

[0123] When the correction direction is a left correction, the correction angle is determined by the following formula :

[0124] ;

[0125] Preferably, in step S1054, correcting the target global path according to the correction angle includes:

[0126] Correcting the coordinates of the collision point based on the correction angle to obtain the coordinates of the correction point.

[0127] Based on the correction point and the points other than the collision point in the target global path, correcting the target global path to obtain the corrected global path.

[0128] The coordinates of the collision point ( , ) are corrected by the following formula:

[0129]

[0130]

[0131] In the formula, Represents the abscissa of the correction point, Represents the ordinate of the correction point.

[0132] It can be understood that the above method for correcting the target global path performs a local correction, so the calculation amount is small and the real-time performance is high.

[0133] Preferably, after correcting the target global path to obtain a corrected global path, the above method further includes:

[0134] Using the corrected global path as the initial global path.

[0135] Based on the initial global path, determining the corresponding smooth path and target global path, and performing collision detection.

[0136] In the case where the collision detection fails, continue to correct until the final target global path passes the collision detection.

[0137] It can be understood that in the case where the final target global path passes the collision detection, it can be ensured that the navigation of the target vehicle based on the above method is safe and effective.

[0138] The embodiment of the present invention also provides a non-transitory machine-readable medium storing a computer program, wherein the computer program is used to cause the computer to execute the method of the embodiment of the present invention when executed by a processor of the computer.

[0139] The embodiment of the present invention also provides a computer program product, including a computer program, wherein the computer program is used to cause a computer to execute the method of the embodiment of the present invention when executed by a processor of the computer.

[0140] The embodiment of the present invention also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program capable of being executed by the at least one processor, and the computer program is used to cause the electronic device to execute the method of the embodiment of the present invention when executed by the at least one processor.

[0141] Referring to Figure 3 , the block diagram of the structure of an electronic device that can be a server or a client of the embodiment of the present invention will now be described, which is an example of a hardware device applicable to various aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computer devices, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0142] As Figure 3As shown, the electronic device includes a computing unit 301, which can execute various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 302 or a computer program loaded from a storage unit 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device can also be stored. The computing unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0143] Multiple components in the electronic device are connected to the I / O interface 305, including: an input unit 306, an output unit 307, a storage unit 308, and a communication unit 309. The input unit 306 can be any type of device capable of inputting information into the electronic device. The input unit 306 can receive input digital or character information, and generate key signal inputs related to the user settings and / or function controls of the electronic device. The output unit 307 can be any type of device capable of presenting information, and can include but is not limited to a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 308 can include but is not limited to a magnetic disk, an optical disk. The communication unit 309 allows the electronic device to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks, and can include but is not limited to a modem, a network card, an infrared communication device, and / or a wireless communication transceiver, such as a Bluetooth device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0144] The computing unit 301 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 301 include but are not limited to a CPU, a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing units, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 301 executes the various methods and processes described above. For example, in some embodiments, the method embodiments of the present invention can be implemented as a computer program, which is tangibly contained in a machine-readable medium, such as the storage unit 308. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device via the ROM 302 and / or the communication unit 309. In some embodiments, the computing unit 301 can be configured to execute the above methods in any other appropriate manner (for example, by means of firmware).

[0145] The computer program for implementing the method of the embodiment of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to the processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, such that when the computer programs are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer programs can be executed entirely on the machine, partially on the machine, executed partially on the machine and partially on a remote machine as an independent software package, or executed entirely on a remote machine or server.

[0146] In the context of the embodiments of the present invention, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable signal medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, or infrared system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0147] It should be noted that the term "including" and its variations used in the embodiments of the present invention are open-ended, that is, "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The modifications of "one" and "plural" mentioned in the embodiments of the present invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise clearly specified in the context, it should be understood as "one or more". The descriptions of the terms "first", "second", etc. are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly specifying the quantity of the indicated technical features.

[0148] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the embodiments of the present invention are all information and data that have been authorized by the user or fully authorized by all parties. Moreover, the collection, use, and processing of the relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions, and corresponding operation entrances are provided for the user to select authorization or rejection.

[0149] In the method embodiments provided by the embodiments of the present invention, the steps described may be executed in different orders and / or executed in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The protection scope of the present invention is not limited in this regard.

[0150] The term "embodiment" in this specification means that the specific features, structures or characteristics described in combination with the embodiments may be included in at least one embodiment of the present invention. The phrase appears in various positions in the specification does not necessarily mean the same embodiment, nor does it mean being independent or alternative to other embodiments and mutually exclusive. The various embodiments in this specification are described in a related manner, and the same or similar parts between the embodiments are referred to each other. In particular, for the device, equipment, and system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts refer to the partial description of the method embodiments.

[0151] The above-described embodiments merely represent several implementation manners of the present invention, and the description thereof is relatively specific and detailed, but should not be construed as a limitation on the protection scope. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the appended claims.

Claims

1. An adaptive path planning method for an autonomous driving vehicle, characterized in that: include: Selecting a prospect point based on the current position of the target vehicle and the pre-planned initial global path; Planning a smooth path according to the current position point and the prospect point; Based on the smooth path, the initial global path is updated to obtain a target global path, wherein the target global path includes a smooth path from the current position point to the prospect point, and an initial global path after the prospect point; Performing collision detection on the target global path; In the case where the collision detection passes, the target vehicle is navigated based on the target global path.

2. The method according to claim 1, characterized in that After navigating the target vehicle based on the target global path, the method further includes: In the next planning cycle, the target global path is used as the initial global path, and replanned to obtain the corresponding target global path; The planning period is adjustable in duration and frequency.

3. The method according to claim 1, characterized in that The method further comprises: If the collision detection fails, determining a collision point of the target global path; Calculating the collision rate of the collision point, and determining the correction direction according to the collision rate; determining a correction angle based on the correction direction; The target global path is corrected according to the correction angle.

4. The method according to claim 3, characterized in that Calculating the collision rate of the collision point and determining the correction direction according to the collision rate include: Calculating the left collision overlap area and the right collision overlap area between the target vehicle and the collision point respectively; Calculating a left collision rate based on the left collision overlap area and vehicle parameters, and calculating a right collision rate based on the right collision overlap area and the vehicle parameters, wherein the vehicle parameters include vehicle length and vehicle width; When the left collision rate is greater than the right collision rate, correct to the right; When the right collision rate is greater than the left collision rate, the vehicle is corrected to the left.

5. The method according to claim 3, characterized in that: Determining a correction angle based on the correction direction includes: Determining a correction angle based on an angle between the first direction and the orientation of the target vehicle, and the correction direction; The first direction is the direction from the current position of the target vehicle to the collision point.

6. The method according to claim 5, characterized in that Correcting the target global path according to the correction angle includes: Correcting the coordinates of the collision point based on the correction angle to obtain the coordinates of the correction point; Based on the correction point and points in the target global path other than the collision point, the target global path is corrected to obtain a corrected global path.

7. The method according to claim 6, characterized in that After correcting the target global path to obtain a corrected global path, the method further includes: Determine a smooth path and a target global path corresponding to the modified global path, and perform collision detection; If the collision detection fails, corrections are continued until the final target global path passes the collision detection.

8. The method according to claim 1, characterized in that: Planning a smooth path according to the current position point of the target vehicle and the prospect point, including: Determine the radius of the arc based on the distance between the prospect point and the current position point, and the angle between a second direction and the orientation of the target vehicle, wherein the second direction is the direction from the current position point to the prospect point; The circular arc is used as the smooth path.

9. The method according to claim 8, characterized in that Navigating the target vehicle based on the target global path includes: calculating a steering angle based on a radius of the smoothed path; The target vehicle is navigated based on the steering angle within a preset time, wherein the preset time is the time for the target vehicle to track the prospect point.

10. The method according to any one of claims 1 to 9, characterized in that Based on the current position of the target vehicle and the pre-planned initial global path, before selecting the prospect point by the sampling planning method, the method further includes: Generate a grid map based on the initial position information, target position information and obstacle information of the target vehicle; An initial global path is planned in the plane rectangular coordinate system of the grid map by using a path planning algorithm.

11. An electronic device, comprising: A processor and a memory storing a program, wherein the program comprises instructions, which, when executed by the processor, cause the processor to perform the method according to any one of claims 1 to 10.

Images