An automatic driving method and apparatus thereof, storage medium, and electronic device

Abstract

The application relates to an automatic driving method and device, a storage medium and an electronic device, which comprises the following steps: acquiring vehicle surrounding environment information at a current time, and generating passable area information at the current time according to the vehicle surrounding environment information at the current time; generating a grid map at the current time according to the passable area information at the current time; the grid map comprises passable cells and impassable cells; acquiring a vehicle position at the current time; performing path searching in the grid map from the position of the vehicle in the grid map at the current time as a starting point to obtain a plurality of passable paths; selecting an optimal passable path from the plurality of passable paths according to a preset path screening strategy, and controlling the vehicle to move according to the optimal passable path; wherein the preset path screening strategy is to make the vehicle travel towards a more spacious area and guide the vehicle to travel towards a preset driving direction.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and more specifically to an autonomous driving method and apparatus, storage medium, and electronic device. Background Technology

[0002] like Figure 1 As shown, when facing narrow roads, the vehicle's driver assistance system uses ultrasonic waves or surround-view cameras to detect the surrounding environment and alert the driver to obstacles through visual and audible warnings to prevent scratches. However, with this system, the driver still needs to manually control the vehicle through narrow sections, requiring a high level of driving skill.

[0003] In addition, such as Figure 2 As shown, similar situations can occur when assisted parking. If the distance between the vehicles on both sides of the parking space is narrow and the driving distance to the left and right front is limited, the driver can only be alerted to the obstacles around the vehicle through visual and audible warnings. The driver needs to control the vehicle to park out of the parking space, which requires a high level of driving skill. Summary of the Invention

[0004] The purpose of this invention is to provide an autonomous driving method and device, storage medium, and electronic equipment to enable autonomous driving through narrow road sections or automatically parking out of narrow parking spaces.

[0005] To achieve the above objectives, embodiments of this application provide an autonomous driving method, the method comprising:

[0006] Obtain the vehicle's surrounding environment information at the current moment, and generate the passable area information at the current moment based on the vehicle's surrounding environment information at the current moment;

[0007] A raster map for the current moment is generated based on the passable area information at the current moment; the raster map includes passable cells and impassable cells;

[0008] Obtain the vehicle's current location, and use the vehicle's current location in the grid map as the starting point to perform a path search in the grid map to obtain multiple travel paths;

[0009] The system selects an optimal travel path from the multiple travel paths according to a preset path selection strategy, and controls vehicle movement based on the optimal travel path. The preset path selection strategy aims to guide the vehicle to travel towards a more open area and to guide the vehicle in a preset travel direction.

[0010] The autonomous driving method provided in this application embodiment obtains the current traversable area information based on the vehicle's surrounding environment information collected by the vehicle-mounted perception module (e.g., camera, ultrasonic sensor, lidar). It then generates a corresponding grid map based on this traversable area information, performs path search based on the grid map to obtain multiple paths, selects the optimal traversable path from these paths according to a pre-set path selection strategy, and controls the vehicle's lateral and longitudinal movement according to the optimal traversable path. The selection principle of the path selection strategy is to guide the vehicle towards a more open area in a narrow space. Simultaneously, during path planning, the vehicle is guided to travel in a preset driving direction, which refers to the target direction, i.e., through which the vehicle travels through narrow road sections or exits parking spaces.

[0011] In some solutions, selecting an optimal travel path from the multiple travel paths according to a preset path filtering strategy specifically includes:

[0012] The area of ​​the trial zone at each moment during the process of the vehicle traveling through each passage path is obtained; wherein, the trial zone is defined as: the vehicle projection area is obtained by projecting the vehicle onto the corresponding vehicle position in the grid map according to the vehicle body parameters and grid map parameters, and the vehicle projection area is expanded to the surrounding areas by a preset threshold distance to obtain the extended area, and the passable area in the extended area is the trial zone.

[0013] Obtain the clockwise angle between the vehicle's driving direction and the preset driving direction between any two adjacent moments during the process of the vehicle traveling through each traffic path; wherein, the vehicle travels a preset unit distance laterally or longitudinally between two adjacent moments;

[0014] Based on the area of ​​the test area at each time moment and the clockwise angle between the vehicle's driving direction and the preset driving direction between any two adjacent times, an optimal driving path is selected from the multiple driving paths.

[0015] In some schemes, selecting an optimal travel path from multiple travel paths based on the area of ​​the trial area at each time moment and the clockwise angle between the vehicle's travel direction and the preset travel direction between any two adjacent time moments specifically includes:

[0016] The path revenue of the multiple travel paths is calculated according to the preset path revenue function R, and the travel path with the largest path revenue is selected as the optimal travel path.

[0017] The path revenue function R is as follows:

[0018]

[0019] Among them, t iLet y(t) be the i-th moment in the process of the vehicle traveling through the passage path, i∈(2,x), and at time t1 the vehicle is located at the starting point. i Let y(t) be the area of ​​the trial region at time i. i-1 Let θ be the area of ​​the trial region at time i-1. i Let l be the clockwise angle between the vehicle's driving direction and the preset driving direction between the (i-1)th time and the ith time, and let C be the preset unit distance. d C m C t All of these are preset constants.

[0020] In some implementations, the preset driving direction is the same as the vehicle gear position when the method is first executed.

[0021] In some solutions, generating a raster map based on the passable area information at the current moment includes:

[0022] Generate a first grid map based on the passable area information at the current moment;

[0023] Obtain the raster map of the historical moment, and expand the map information of the first raster map based on the raster map of the historical moment to obtain the raster map of the current moment.

[0024] In some solutions, acquiring information about the vehicle's surrounding environment and generating passable area information based on that information specifically includes:

[0025] Acquire surrounding environment images captured by vehicle-mounted cameras, and obtain first passable area information based on vision according to the surrounding environment images;

[0026] Acquire ultrasonic data collected by an on-board ultrasonic sensor, and obtain second passable area information based on the ultrasonic data.

[0027] Acquire laser point cloud data collected by vehicle-mounted lidar, and obtain third passable area information based on the laser point cloud data;

[0028] The first passable area information, the second passable area information, and the third passable area information are merged to generate passable area information.

[0029] In some solutions, the process of fusing the first accessible area information, the second accessible area information, and the third accessible area information to generate accessible area information specifically includes:

[0030] Obtain preset weight parameters for the first passable area information, the second passable area information, and the third passable area information within multiple detection distance ranges; wherein, the preset weight parameters for the first passable area information, the second passable area information, and the third passable area information are different within multiple detection distance ranges;

[0031] According to the preset weight parameters, the first passable area information, the second passable area information, and the third passable area information within each detection distance range are weighted and fused to obtain the passable area information within each detection distance range;

[0032] The final passable area information is obtained by analyzing passable area information from multiple detection distance ranges.

[0033] The plurality of detection distance ranges include: a distance range beyond the detection range of the vehicle-mounted camera and the vehicle-mounted ultrasonic sensor; a distance range that can be detected by the vehicle-mounted camera, the vehicle-mounted ultrasonic sensor and the vehicle-mounted LiDAR; and a distance range that can be detected by the vehicle-mounted camera and the vehicle-mounted ultrasonic sensor but cannot be detected by the vehicle-mounted LiDAR.

[0034] This application also provides an autonomous driving device, including a module for performing the autonomous driving method described above.

[0035] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the autonomous driving method described above.

[0036] This application also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the autonomous driving method described above. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings required in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a schematic diagram illustrating a scenario where traffic is passing through a narrow road.

[0039] Figure 2 A schematic diagram illustrating a scenario where parking is done in a complex parking space.

[0040] Figure 3This is a flowchart of an autonomous driving method according to one embodiment of the present invention.

[0041] Figure 4 This is a schematic diagram illustrating the path search principle in one embodiment of the present invention.

[0042] Figure 5 This is a schematic diagram of the vehicle projection area and the test area in one embodiment of the present invention.

[0043] Figure 6 This is a schematic diagram of a grid map expansion in one embodiment of the present invention.

[0044] Figure 7 This is a schematic diagram of an autonomous driving device according to one embodiment of the present invention. Detailed Implementation

[0045] The detailed description of the accompanying drawings is intended to illustrate the presently preferred embodiments of the invention and is not intended to represent only the forms in which the invention can be implemented. It should be understood that the same or equivalent functions can be accomplished by different embodiments intended to be included within the spirit and scope of the invention.

[0046] One embodiment of the present invention provides an autonomous driving method, see below. Figure 3 The method in this embodiment includes the following steps:

[0047] Step S1: Obtain the vehicle's surrounding environment information at the current moment, and generate the passable area information at the current moment based on the vehicle's surrounding environment information at the current moment.

[0048] Specifically, the passable area information refers to information about roads / areas that the vehicle can pass through. The vehicle's surrounding environment information can be detected by the onboard perception module (e.g., camera, ultrasonic sensor, lidar) and sent to the CAN / CANFD bus. In step S1, the vehicle's surrounding environment information at the current moment can be obtained from the CAN bus. Based on the vehicle's surrounding environment information, obstacles (e.g., pedestrians, vehicles), roads, and other information in the vehicle's surrounding environment can be identified, and then the corresponding passable area information can be generated. In one example, a pre-trained neural network model is set up, and the vehicle's surrounding environment information is input into the neural network model to identify the passable area and obtain the passable area information at the current moment.

[0049] Step S2: Generate a raster map for the current time based on the passable area information at the current time; the raster map includes passable cells and impassable cells.

[0050] Specifically, the raster map consists of multiple cells arranged in rows and columns. Each cell represents a small area. Different values ​​can be assigned to each cell to indicate the accessibility status of the small area corresponding to the cell, i.e., whether it is accessible or inaccessible. The characteristics of the raster map are that it is easy to build, represent and save, and it is very convenient for planning short paths. Therefore, this embodiment uses the raster map to quickly determine the access path.

[0051] Step S3: Obtain the vehicle's current location, and use the vehicle's current location in the grid map as the starting point to perform a path search in the grid map to obtain multiple travel paths.

[0052] Specifically, in step S3, a path search can be performed according to a preset path search algorithm. In this embodiment, the path search algorithm is not limited to one type, for example:

[0053] See Figure 4 At time t1, the vehicle's position in the grid map is at the starting point p0, and based on this starting point, it moves in three preset directions at time t1. Figure 4 The vehicle makes its first exploratory move (indicated by the middle arrow: left, forward, right). At time t2, three new vehicle positions p1, p2, and p3 are obtained. Based on these vehicle positions p1, p2, and p3, the vehicle makes its first exploratory move in multiple preset directions at time t2. At time t3, the vehicle positions p4, p5, p6, p7, p8, p9, and p1 are obtained. 10 Multiple new vehicle location points are obtained through this process. Each exploratory movement yields many new vehicle location points. The line connecting the two vehicle location points before and after each movement forms the vehicle's trajectory, resulting in multiple travel paths. For example, points p0, p2, and p8 form one travel path, and points p0, p3, and p8 also form another. It should be noted that the number of exploratory movements before generating a travel path can be set according to actual technical requirements, allowing for simultaneous lateral and longitudinal vehicle movement and path searching.

[0054] Step S4: Select an optimal travel path from the multiple travel paths according to a preset path selection strategy, and control the vehicle movement according to the optimal travel path; wherein, the preset path selection strategy is to make the vehicle travel towards a more open area and guide the vehicle to travel in a preset direction.

[0055] Specifically, the autonomous driving method provided in this embodiment obtains the current passable area information based on the vehicle's surrounding environment information collected by the vehicle-mounted perception module (e.g., camera, ultrasonic sensor, lidar). It then generates a corresponding grid map based on this passable area information, performs path search based on the grid map to obtain multiple paths, selects the optimal path from these paths according to a pre-set path selection strategy, and controls the vehicle's lateral and longitudinal movement according to the optimal path. The selection principle of the path selection strategy is to guide the vehicle towards a more open area in a narrow space. Simultaneously, during path planning, the vehicle is guided to travel in a preset driving direction, which refers to the target direction, i.e., through which the vehicle travels through narrow road sections or exits parking spaces. The preset driving direction is consistent with the vehicle's gear position direction at the start of the method's execution.

[0056] It should be noted that, due to the confined space, the vehicle moves at a low speed during the control of its lateral and longitudinal movements. The grid map is continuously updated according to step S2, and the path search in step S3 and the path filtering in step S4 are performed again. If a new obstacle is detected in the next area in the direction of travel, emergency braking is applied to avoid a collision.

[0057] In some embodiments, step S4 specifically includes:

[0058] Step S41: Obtain the area of ​​the trial region at each moment during the vehicle's journey along each travel path; where, for example... Figure 5 As shown, the test area is defined as follows: the vehicle projection area is obtained by projecting the vehicle onto the corresponding vehicle position in the grid map according to the vehicle body parameters and grid map parameters. The vehicle projection area is expanded to the surrounding areas by a preset threshold distance to obtain the extended area. The passable area in the extended area is the test area.

[0059] Specifically, the process of a vehicle traveling through each route takes a certain amount of time, which includes multiple moments. In this embodiment, the area of ​​the trial area at each moment is used to determine whether the area in the direction of the route is more open. For example, assuming the vehicle body is 5 meters long and 2.5 meters wide, and one cell represents a 5-meter by 5-meter area, then the projection of the vehicle in the grid map is 1 / 2 of the cell.

[0060] Step S42: Obtain the clockwise angle between the vehicle's driving direction and the preset driving direction between any two adjacent moments during the process of the vehicle traveling through each traffic path; wherein, the vehicle travels a preset unit distance laterally or longitudinally between two adjacent moments.

[0061] Specifically, such as Figure 4In the path search process shown, the horizontal direction is to move left or right according to the vehicle's current orientation, and the vertical direction is to move forward according to the vehicle's current orientation. The left, right, and forward directions are the vehicle's travel directions between any two adjacent moments.

[0062] Step S43: Select an optimal travel path from the multiple travel paths based on the area of ​​the test area at each time moment and the clockwise angle between the vehicle travel direction and the preset travel direction between any two adjacent time moments.

[0063] Specifically, the area of ​​the test area at each moment can be used to determine whether the vehicle is moving towards a more open area. The clockwise angle between the vehicle's driving direction and the preset driving direction between any two adjacent moments can be used to determine whether the vehicle is moving towards the preset driving direction. By comprehensively considering the area of ​​the test area and the clockwise angle, an optimal travel path is selected so that the vehicle moves towards a more open area and is guided to move towards the preset driving direction.

[0064] In some embodiments, selecting an optimal travel path from the plurality of travel paths based on the area of ​​the trial area at each time moment and the clockwise angle between the vehicle's travel direction and the preset travel direction between any two adjacent time moments specifically includes:

[0065] The path revenue of the multiple travel paths is calculated according to the preset path revenue function R, and the travel path with the largest path revenue is selected as the optimal travel path.

[0066] The path revenue function R is as follows:

[0067]

[0068] Among them, t i Let y(t) be the i-th moment in the process of the vehicle traveling through the travel path, i∈(2,x), where x is the last moment. The value of x depends on the length of the path and the travel time. At moment t1, the vehicle is located at the starting point. i Let y(t) be the area of ​​the trial region at time i. i-1 Let θ be the area of ​​the trial region at time i-1. i Let l be the clockwise angle between the vehicle's driving direction and the preset driving direction between the (i-1)th time and the ith time, and let C be the preset unit distance. d C m C t All are preset constants;

[0069] Specifically, in function R This means that by monitoring the changes in the sum of the areas occupied by the grid map in the test area towards a larger direction, the reward design guides vehicles to drive towards more open areas; This indicates that the vehicle can consider moving in multiple directions, such as forward, backward, left, and right, and the reward design guides the vehicle to keep moving in the preset driving direction; C t As a time-dependent constant, considering efficiency, the longer the shift time, the more efficient C becomes. t The larger the value set, the greater the corresponding negative return.

[0070] In some embodiments, step S2 specifically includes:

[0071] Generate a first grid map based on the passable area information at the current moment;

[0072] Obtain the raster map of the historical moment, and expand the map information of the first raster map based on the raster map of the historical moment to obtain the raster map of the current moment.

[0073] Specifically, since the detection range of the vehicle-mounted sensor unit is limited, the generated grid map also has a limited area. However, as the vehicle moves, the actual detection area of ​​the vehicle-mounted sensor unit will change, such as... Figure 6 As shown, by adding map information from the raster map at a historical moment to the first raster map, the map information of the raster map at the current moment can be expanded and enriched.

[0074] In some embodiments, step S1 specifically includes:

[0075] Step S11: Obtain the surrounding environment image captured by the vehicle-mounted camera, and obtain the first passable area information based on the vision based on the surrounding environment image;

[0076] Step S12: Obtain ultrasonic data collected by the vehicle-mounted ultrasonic sensor, and obtain second passable area information based on the ultrasonic data.

[0077] Step S13: Obtain laser point cloud data collected by vehicle-mounted lidar, and obtain third passable area information based on the point cloud data.

[0078] Step S14: The first passable area information, the second passable area information, and the third passable area information are merged to generate passable area information.

[0079] Specifically, the method in this embodiment solves the limitations of insufficient ranging accuracy of a single camera and susceptibility to light, as well as the inability of a single ultrasonic sensor to classify targets and the possibility of misidentification within the passable range, by fusing data from a camera, ultrasonic sensor, and lidar. By adding lidar detection, it can achieve ultra-long-distance environmental model perception and centimeter-level distance perception, making the generated passable area information more accurate and reliable.

[0080] In some embodiments, step S14 specifically includes:

[0081] Step S141: Obtain preset weight parameters for the first passable area information, the second passable area information, and the third passable area information within multiple detection distance ranges;

[0082] Step S142: According to the preset weight parameters, the first passable area information, the second passable area information, and the third passable area information within each detection distance range are weighted and fused to obtain the passable area information within each detection distance range;

[0083] Step S143: Obtain the final passable area information by combining the passable area information from multiple detection distance ranges;

[0084] The plurality of detection distance ranges include: a distance range beyond which neither the vehicle-mounted camera nor the vehicle-mounted ultrasonic sensor can detect; a distance range that can be detected by the vehicle-mounted camera, the vehicle-mounted ultrasonic sensor, and the vehicle-mounted LiDAR; and a distance range that can be detected by both the vehicle-mounted camera and the vehicle-mounted ultrasonic sensor but cannot be detected by the vehicle-mounted LiDAR.

[0085] Specifically, this embodiment fuses the passable horizontal areas output by the camera, ultrasonic sensor, and lidar. For example, by using a Bayesian method to adjust the sensor weights at different locations, a grid map of the vehicle's surrounding environment is generated.

[0086] Because the detection ranges of cameras, ultrasonic sensors, and lidar differ, specifically, based on practical considerations, we assume that the detection range of a typical surround-view camera is within 10m, ultrasonic sensors within 5m, and lidar within 100m. At any given moment, different strategies and weights are used to construct a grid map of the environment at different distances around the vehicle, generating multi-sensor fusion information for the communicable area. Since lidar has a long detection range, in distances beyond the detection range of cameras and ultrasonic sensors, only the third drivable area information from lidar is used to construct the grid map. Within the detection range of cameras, ultrasonic sensors, and lidar, due to the inaccuracy of camera ranging, the drivable area distance information is primarily based on the fusion of the second and third drivable area information, while the obstacle type corresponding to the drivable area is primarily based on the fusion of the first and third drivable area information. Within the detection range of both cameras and ultrasonic sensors but not of lidar, the drivable area distance information is primarily based on the second drivable area information, while the obstacle type corresponding to the drivable area is primarily based on the first drivable area information.

[0087] For example, within the range that can be detected by vehicle-mounted cameras, vehicle-mounted ultrasonic sensors, and vehicle-mounted lidar, the weighting coefficient for the first drivable area information is set to 0.3, the weighting coefficient for the second drivable area information is set to 0.7, and the weighting coefficient for the third drivable area information is set to 1.

[0088] In addition, since the lidar point cloud data contains precise distance and height information, the height information is mainly based on the lidar detection data. In step 13, the height information provided in the third passable area information is compared with its own height information to eliminate intersections that are actually impassable during the process of eliminating passable areas.

[0089] In some embodiments, step S4, controlling vehicle movement according to the travel path, specifically includes:

[0090] If, during movement, the vehicle's camera, ultrasonic sensor, or lidar detects a new obstacle on the travel path, the vehicle will be stopped to avoid scraping or colliding with the obstacle.

[0091] In some embodiments, step S4, controlling vehicle movement according to the travel path, further includes:

[0092] If there are no obstacles within the preset distance range of the travel path in the grid map, or if the driver receives a braking command during autonomous driving and the driver selects to interrupt the autonomous driving function, or if the preset autonomous driving time is exceeded, the vehicle will be stopped and the driver will be allowed to take over the vehicle.

[0093] As described in the above embodiments, the method of this embodiment combines camera, ultrasonic, and lidar sensors in terms of perception, solving the problem of environmental perception limitations caused by physical factors for single-type perception sensors. Simultaneously, it leverages the detection range advantage of lidar to perceive environmental conditions and improve recognition accuracy at a greater distance beyond the driving direction, enhancing its applicability to narrow road passages or parking in confined spaces. Furthermore, through effective path planning cost function design, it provides a solution for vehicles to autonomously navigate complex narrow roads or complex parking spaces, expanding the application scenarios that autonomous driving functions can cover.

[0094] Another embodiment of this application also provides an autonomous driving device, including a module for performing the autonomous driving method described in the above embodiments;

[0095] See Figure 7 The module described in this embodiment includes:

[0096] The passable area acquisition module 1 is used to acquire the vehicle's surrounding environment information at the current moment, and generate passable area information at the current moment based on the vehicle's surrounding environment information at the current moment.

[0097] Map generation module 2 is used to generate a raster map for the current time based on the passable area information at the current time; the raster map includes passable cells and impassable cells;

[0098] The path search module 3 is used to obtain the vehicle's current position and perform path search in the grid map using the vehicle's current position in the grid map as the starting point to obtain multiple travel paths.

[0099] The route selection module 4 is used to select an optimal route from the multiple routes according to a preset route selection strategy, and control the vehicle movement according to the optimal route; wherein, the preset route selection strategy is to make the vehicle travel towards a more open area and guide the vehicle to travel in a preset direction.

[0100] The autonomous driving devices described in the embodiments above are merely illustrative. The modules described as separate components may or may not be physically separate. The components of a module may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of the autonomous driving device solution in the embodiments, depending on actual needs.

[0101] It should be noted that the autonomous driving device in the above embodiments corresponds to the autonomous driving method in the above embodiments. Therefore, the parts of the autonomous driving device in the above embodiments that are not described in detail can be obtained by referring to the content of the autonomous driving method in the above embodiments. That is, the specific steps recorded in the autonomous driving method in the above embodiments can be understood as the functions that the autonomous driving device in the above embodiments can achieve, and will not be described again here.

[0102] Furthermore, if the autonomous driving device of the above embodiments is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0103] Another embodiment of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the autonomous driving method as described in the above embodiments.

[0104] Specifically, the computer-readable storage medium may include any entity or recording medium capable of carrying the computer program instructions, such as a USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media.

[0105] Another embodiment of the present invention provides an electronic device including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the autonomous driving method described in the above embodiments.

[0106] The electronic device may also include a bus connecting different components, including memory and processor. The memory may include a computer-readable medium in the form of volatile memory, such as random access memory (RAM) and / or cache memory. The memory may also include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of this application. The electronic device may also communicate with one or more external devices (e.g., keyboard, pointing device, display, etc.), and with one or more devices that enable a user to interact with the electronic device, and / or with any device (e.g., a network interface card) that enables the electronic device to communicate with one or more other computing devices, such communication may be performed via an input / output (I / O) interface, and the electronic device may also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via a network adapter.

[0107] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. An autonomous driving method, characterized in that, include: Obtain the vehicle's surrounding environment information at the current moment, and generate the passable area information at the current moment based on the vehicle's surrounding environment information at the current moment; Generate a raster map of the current time based on the passable area information at the current time; The raster map includes passable cells and impassable cells; Obtain the vehicle's current location, and use the vehicle's current location in the grid map as the starting point to perform a path search in the grid map to obtain multiple travel paths; The system selects an optimal travel path from the multiple travel paths according to a preset path selection strategy, and controls vehicle movement based on this optimal path. The preset path selection strategy directs the vehicle towards a more open area and guides it in a preset direction. Specifically, selecting the optimal path from the multiple travel paths according to the preset path selection strategy includes: The area of ​​the trial zone at each moment during the process of the vehicle traveling through each passage path is obtained; wherein, the trial zone is defined as: the vehicle projection area is obtained by projecting the vehicle onto the corresponding vehicle position in the grid map according to the vehicle body parameters and grid map parameters, and the vehicle projection area is expanded to the surrounding areas by a preset threshold distance to obtain the extended area, and the passable area in the extended area is the trial zone. Obtain the clockwise angle between the vehicle's driving direction and the preset driving direction between any two adjacent moments during the process of the vehicle traveling through each traffic path; wherein, the vehicle travels a preset unit distance laterally or longitudinally between two adjacent moments; Based on the area of ​​the test area at each time moment and the clockwise angle between the vehicle's driving direction and the preset driving direction between any two adjacent times, an optimal driving path is selected from the multiple driving paths.

2. The method according to claim 1, characterized in that, The step of selecting an optimal travel path from the multiple travel paths based on the area of ​​the test area at each moment and the clockwise angle between the vehicle's travel direction and the preset travel direction between any two adjacent moments specifically includes: The path revenue of the multiple travel paths is calculated according to the preset path revenue function R, and the travel path with the largest path revenue is selected as the optimal travel path. The path revenue function R is as follows: Among them, t i Let t1 be the i-th moment in the process of the vehicle traveling through the traffic path, where i ∈ (2, x), and the vehicle is located at the starting point at time t1. Let be the area of ​​the trial region at time i. Let be the area of ​​the trial region at time i-1. The clockwise angle between the vehicle's driving direction and the preset driving direction is defined between the (i-1)th time and the ith time. For preset unit distance, , , All of these are preset constants.

3. The method according to claim 1, characterized in that, The preset driving direction is the same as the vehicle gear direction when the steps of the method are started.

4. The method according to claim 1, characterized in that, The step of generating a raster map for the current time based on the passable area information at the current time includes: Generate a first grid map based on the passable area information at the current moment; Obtain the raster map of the historical moment, and expand the map information of the first raster map based on the raster map of the historical moment to obtain the raster map of the current moment.

5. The method according to claim 1, characterized in that, The step of acquiring information about the vehicle's surrounding environment and generating passable area information based on that information specifically includes: Acquire surrounding environment images captured by vehicle-mounted cameras, and obtain first passable area information based on vision according to the surrounding environment images; Acquire ultrasonic data collected by an on-board ultrasonic sensor, and obtain second passable area information based on the ultrasonic data. Acquire laser point cloud data collected by vehicle-mounted lidar, and obtain third passable area information based on the laser point cloud data; The first passable area information, the second passable area information, and the third passable area information are merged to generate passable area information.

6. The method according to claim 5, characterized in that, The step of fusing the first passable area information, the second passable area information, and the third passable area information to generate passable area information specifically includes: Obtain preset weight parameters for the first passable area information, the second passable area information, and the third passable area information within multiple detection distance ranges; wherein, the preset weight parameters for the first passable area information, the second passable area information, and the third passable area information are different within multiple detection distance ranges; According to the preset weight parameters, the first passable area information, the second passable area information, and the third passable area information within each detection distance range are weighted and fused to obtain the passable area information within each detection distance range; The final passable area information is obtained based on passable area information within multiple detection distance ranges; The plurality of detection distance ranges include: a distance range beyond the detection range of the vehicle-mounted camera and the vehicle-mounted ultrasonic sensor; a distance range that can be detected by the vehicle-mounted camera, the vehicle-mounted ultrasonic sensor and the vehicle-mounted LiDAR; and a distance range that can be detected by the vehicle-mounted camera and the vehicle-mounted ultrasonic sensor but cannot be detected by the vehicle-mounted LiDAR.

7. An autonomous driving device, characterized in that, It includes a module for performing the autonomous driving method according to any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the autonomous driving method as described in any one of claims 1 to 6.

9. An electronic device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the autonomous driving method as described in any one of claims 1 to 6.

Images