Path planning method and device, vehicle, and storage medium

Abstract

The application discloses a path planning method and device, a vehicle and a storage medium. The method comprises the following steps: acquiring road edge data and travel trajectory data collected in a path learning process by a vehicle; determining a road boundary line corresponding to a straight road section in the path learning process and a straight planning path of the vehicle corresponding to the straight road section based on the road edge data and the travel trajectory data; determining a turning planning path of the vehicle corresponding to a turning road section in the path learning process based on the road boundary line and the straight planning path; and determining a driving path of the vehicle in the driving process based on the straight planning path and the turning planning path. The data collected by the vehicle in the path learning process is used to determine the straight planning path and the turning planning path corresponding to the straight road section and the turning road section respectively, and then the driving path of the vehicle is obtained, so that the driving path is more in line with the actual environment of the road.

Description

Technical Field

[0001] This application relates to the field of vehicle control, and more specifically, to a path planning method, apparatus, vehicle, and storage medium. Background Technology

[0002] In recent years, as autonomous driving technology has matured, many vehicles have begun to support autonomous driving in known environments. This means that a vehicle can learn a route for a specific road segment beforehand through manual driving. If the vehicle subsequently passes through the same road segment, the user can switch to autonomous driving mode, and the vehicle can then avoid obstacles based on its previous trajectory. However, the autonomous driving trajectory in this scenario may have flaws. For example, if the vehicle avoids a dynamic obstacle while learning its route, but that obstacle disappears in the subsequent autonomous driving phase, the vehicle will typically continue along the learned detour path. This means the autonomous driving path no longer reflects the actual road conditions, thus affecting the user's experience with autonomous driving. Summary of the Invention

[0003] In view of the above problems, this application proposes a route planning method, device, vehicle and storage medium to obtain a driving route that is more consistent with the actual road environment.

[0004] In a first aspect, embodiments of this application provide a path planning method, the method comprising: acquiring road edge data and travel trajectory data collected during a path learning process by a vehicle; determining, based on the road edge data and travel trajectory data, a road boundary line corresponding to a straight section in the path learning process and a straight-line planned path for the vehicle corresponding to the straight section; determining, based on the road boundary line and the straight-line planned path, a turning planned path for the vehicle corresponding to a turning section in the path learning process; and determining, based on the straight-line planned path and the turning planned path, the vehicle's driving path during the driving process.

[0005] Secondly, embodiments of this application provide a path planning device, the device comprising: a data acquisition module, a straight path module, a turning path module, and a driving path module, wherein the data acquisition module is used for the vehicle to acquire road edge data and travel trajectory data collected during the path learning process; the straight path module is used to determine the road boundary line corresponding to the straight section in the path learning process and the straight planned path of the vehicle corresponding to the straight section based on the road edge data and travel trajectory data; the turning path module is used to determine the turning planned path of the vehicle corresponding to the turning section in the path learning process based on the road boundary line and the straight planned path; and the driving path module is used to determine the driving path of the vehicle in the driving process based on the straight planned path and the turning planned path.

[0006] Thirdly, embodiments of this application provide a vehicle, including: one or more processors; a memory; and one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to perform the path planning method provided in the first aspect above.

[0007] Fourthly, embodiments of this application provide a computer-readable storage medium storing program code, which can be invoked by a processor to execute the path planning method provided in the first aspect above.

[0008] The solution provided in this application acquires road edge data and travel trajectory data during the vehicle's path learning process. Based on the road edge data and travel trajectory data, it determines the road boundary line corresponding to the straight-ahead segment and the planned straight-ahead path for the vehicle on that segment. Based on the road boundary line and the planned straight-ahead path, it determines the planned turning path for the vehicle on the turning segment. Based on the planned straight-ahead path and the planned turning path, it determines the vehicle's driving path during the entire journey. By using the data collected during the path learning process, the solution determines the planned straight-ahead path and the planned turning path for the straight-ahead and turning segments, respectively, thereby obtaining the vehicle's driving path and making it more consistent with the actual road environment. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A flowchart illustrating a path planning method provided in one embodiment of this application is shown.

[0011] Figure 2 The diagram illustrates the driving path of the vehicle under two different states in an embodiment of this application.

[0012] Figure 3 A flowchart illustrating a path planning method provided in another embodiment of this application is shown.

[0013] Figure 4 A schematic diagram of the specific process of step S230 in another embodiment of this application is shown.

[0014] Figure 5 A schematic diagram illustrating the principle of determining the target travel point in an embodiment of this application is shown.

[0015] Figure 6 A schematic diagram of the specific process of step S250 in another embodiment of this application is shown.

[0016] Figure 7 A schematic diagram showing the connection of two adjacent straight-line planning paths in an embodiment of this application is shown.

[0017] Figure 8 A schematic diagram of the specific process of step S260 in another embodiment of this application is shown.

[0018] Figure 9 A schematic diagram of the planned turning path corresponding to the turning section in an embodiment of this application is shown.

[0019] Figure 10 A schematic diagram of the structure of a path planning device provided in another embodiment of this application is shown.

[0020] Figure 11 A structural block diagram of a computer device provided in an embodiment of this application is shown.

[0021] Figure 12 A structural block diagram of a computer-readable storage medium provided in an embodiment of this application is shown. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0023] The inventors have proposed a path planning method, device, vehicle, and storage medium provided in the embodiments of this application. By using the data collected by the vehicle during the path learning process, the straight-going plan path and the turning plan path corresponding to the straight-going section and the turning section of the vehicle during the path learning process are determined respectively, thereby obtaining the vehicle's driving path, making the driving path more consistent with the actual road environment.

[0024] The path planning method provided in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0025] Please see Figure 1 , Figure 1 This paper illustrates a flowchart of a path planning method provided in one embodiment of this application. The following will focus on... Figure 1 The process shown is described in detail. The path planning method may specifically include the following steps:

[0026] Step S110: The vehicle acquires road edge data and travel trajectory data collected during the path learning process.

[0027] In this embodiment, if a user requires the vehicle to operate autonomously on a specific road segment, the user must first manually drive the vehicle from the starting point to the destination on that segment. This allows the vehicle to collect road edge data and trajectory data during path learning. Based on this data, the vehicle can determine its driving path for autonomous driving on the specific road segment. This allows the vehicle to prompt the user to initiate autonomous driving when it detects entering the segment, and then proceed according to the predetermined path once in autonomous driving mode. Therefore, to obtain the driving path for autonomous driving on a specific road segment, the vehicle can acquire road edge data and trajectory data from the path learning process when driving the same segment manually.

[0028] Understandably, the trajectory of a vehicle under manual driving conditions can be any curved line. For example... Figure 2 As shown, the curved trajectory represents the vehicle's path through this road segment under manual driving conditions. During path learning, the vehicle collects road edge data and trajectory data to generate the corresponding driving path for that road segment, such as... Figure 2 The straight-line trajectory in the image shows that the driving path generated by the vehicle based on road edge data and travel trajectory data can reduce unnecessary detours during the vehicle's path learning process. Therefore, it can be used as the driving trajectory of the vehicle in autonomous driving mode.

[0029] The vehicle path learning process refers to the path learning process when a vehicle passes through a certain road segment under manual driving conditions. The range of this road segment can be automatically generated based on the user's driving habits, or it can be preset by the user based on their own needs. For example, when the vehicle passes through a certain location, the user can set the vehicle to path learning mode through the human-computer interaction interface. Then the vehicle can automatically collect road edge data and travel trajectory data from the current location and end data collection at the user-defined location.

[0030] In some implementations, vehicles can acquire road edge data during the path learning process using sensors such as onboard cameras, millimeter-wave radar, or ultrasonic radar. In this case, the road edge data acquired by the vehicle through these sensors is in a world coordinate system. The vehicle can convert this world coordinate system road edge data into road coordinate system / SL coordinate system data to facilitate subsequent determination of the vehicle's driving path based on the road edge data.

[0031] In some implementations, if the user requires the vehicle to achieve autonomous driving in areas with weak signals, such as underground parking lots, it may be difficult for the vehicle to collect trajectory data during the path learning process via the Global Navigation Satellite System (GNSS). In this case, the vehicle can acquire trajectory data during the path learning process through its onboard Inertial Measurement Unit (IMU). The IMU sensor only records data such as angular velocity and acceleration along various axes during path learning, but it cannot directly record the coordinate data of the vehicle at various points during the path learning process. After acquiring the angular velocity and angular velocity data at various points during the path learning process, this data can be processed to obtain the coordinate data of each point. Specifically, the vehicle can first acquire the coordinate data of the first point in a certain path learning segment as an initial value. The coordinate data of subsequent points during the path learning process are calculated based on this initial value, using the angular velocity and angular velocity data of that point along various axes acquired by the IMU sensor. It is understandable that the coordinate data of various points in the vehicle's path learning process, obtained from IMU sensors, may contain errors compared to the actual coordinates of those points. This is because the calculation of coordinate data inevitably involves errors, and the coordinates of subsequent points are calculated based on the coordinates of previous points using IMU sensor measurements. Therefore, the calculation errors of the previous points will accumulate in the coordinate data of subsequent points. The longer the vehicle's trajectory, the greater the accumulated errors in the acquired trajectory data. Therefore, the vehicle can acquire a new initial IMU value at regular intervals or after certain distances. The coordinates of subsequent points are then calculated based on the coordinates corresponding to this initial value, thereby eliminating the accumulated errors in the trajectory data prior to the initial value.

[0032] In some implementations, the road edge data and trajectory data collected by the vehicle through different sensors during path learning may not be uniformly distributed. For example, the vehicle may acquire road edge data during path learning using LiDAR at a fixed frequency of 30 frames per second. However, the vehicle's speed during path learning is not uniform, resulting in unevenly distributed road edge data. Similarly, the trajectory data acquired by the vehicle may not be uniformly distributed. Therefore, to facilitate subsequent determination of the vehicle's path based on the road edge and trajectory data, after collecting the road edge and trajectory data during path learning, the vehicle can interpolate these raw data at equal intervals. For example, the vehicle can pre-set the interpolation distance to one meter. After acquiring the raw road edge and trajectory data directly collected by the sensors, the vehicle can filter this raw data, retaining one sampling point every meter. If no data exists at the sampling point location in the raw data, it can determine the location based on the nearest raw data to that sampling point. Thus, the vehicle can obtain uniformly distributed road edge data and travel trajectory data. Furthermore, since the intervals between the road edge data and the travel trajectory data are the same, there can be a correspondence between the road edge data and the travel trajectory data collected by the vehicle.

[0033] Step S120: Based on the road edge data and travel trajectory data, determine the road boundary line corresponding to the straight road segment in the path learning process and the straight planning path of the vehicle corresponding to the straight road segment.

[0034] In this embodiment, the driving path of a vehicle in autonomous driving mode can include a straight-line planned path for a straight-line segment and a turning planned path for a turning segment. In other words, the vehicle can directly divide the driving process into straight-line and turning segments according to the direction of travel, and determine the straight-line planned path for the straight-line segment and the turning planned path for the turning segment based on the road edge data and trajectory data collected during path learning. Specifically, after acquiring the road edge data and trajectory data during path learning, the vehicle can first determine the straight-line planned path for the straight-line segment based on this data, and then determine the turning planned path connecting the two straight-line planned paths based on the straight-line planned path. The trajectory data acquired by the vehicle can include the direction data acquired during path learning. Therefore, the vehicle can divide the trajectory data during path learning into trajectory data for the straight-line segment and trajectory data for the turning segment based on the direction data in the trajectory data, and then obtain the straight-line planned path for the straight-line segment based on the trajectory data for the straight-line segment.

[0035] Step S130: Based on the road boundary line and the straight-line planning path, determine the turning planning path corresponding to the turning segment of the vehicle in the path learning process.

[0036] In this embodiment, after determining the road boundary line and straight-line planning path corresponding to the straight-line segment based on road edge data and travel trajectory data, the vehicle can further determine the turning planning path between two adjacent straight-line planning paths based on the straight-line planning path. It should be understood that the driving path determined in this application is a reference path for the vehicle under autonomous driving conditions. Furthermore, since there are inherent limitations in autonomous driving mode, such as the steering wheel angle needing to be less than a preset threshold (meaning the steering wheel can only turn within a smaller range than when manually driving), the turning path in the travel trajectory data obtained during data collection may not be directly used as the turning planning path for the turning segment. This is because the vehicle is manually driven during data collection, and there is no limit to the steering wheel angle at this time. Therefore, the steering wheel angle when the vehicle passes through the turning segment may exceed the limit for steering wheel angle under autonomous driving conditions, making it impossible for the vehicle to follow the travel trajectory completely under autonomous driving. Therefore, for the turning planning path corresponding to the turning segment, the vehicle can obtain it through the straight planning path corresponding to the two adjacent straight segments, without having to refer to the road edge data and travel trajectory data collected during the path learning process.

[0037] Step S140: Based on the straight-line planning path and the turning planning path, determine the driving path of the vehicle during the path learning process.

[0038] In this embodiment, after the vehicle determines the straight-line and turning planned paths during the path learning process, all the straight-line and turning planned paths in the path learning process can be connected to obtain the vehicle's driving path during the path learning process. This driving path is the driving path of the vehicle during autonomous driving. It is worth noting that the driving path obtained at this time is only a reference driving path for the vehicle in autonomous driving mode. That is to say, the vehicle will not drive exactly according to the driving path obtained at this time during autonomous driving. Instead, it will perform obstacle avoidance driving based on the road condition information obtained by sensors in real time. Because the road conditions during the vehicle's path learning process are inevitably more complex, the driving path determined at this time cannot include dynamic obstacles that may appear on the roadside during the actual driving process. Therefore, it is impossible to instruct the vehicle to detour around these obstacles in the driving path. Therefore, the vehicle can only actively detour around obstacles in the road based on real-time environmental data during the actual driving process.

[0039] The path planning method provided in this application involves acquiring road edge data and travel trajectory data collected during the vehicle's path learning process. Based on the road edge data and travel trajectory data, the method determines the road boundary line corresponding to the straight-ahead segment and the planned straight-ahead path for the vehicle on that segment. Based on the road boundary line and the planned straight-ahead path, the method determines the planned turning path for the vehicle on the turning segment. Based on the planned straight-ahead path and the planned turning path, the method determines the vehicle's actual driving path. By using the data collected during the path learning process, the method determines the planned straight-ahead and planned turning paths for the straight-ahead and turning segments, respectively, thereby obtaining the vehicle's driving path and making it more consistent with the actual road environment.

[0040] Please see Figure 3 , Figure 3 A flowchart illustrating a path planning method provided in another embodiment of this application is shown below. Figure 3 The process shown is described in detail. The path planning method may specifically include the following steps:

[0041] Step S210: The vehicle acquires road edge data and travel trajectory data collected during the path learning process.

[0042] In this embodiment, step S210 can be referred to the content of other embodiments, and will not be repeated here.

[0043] Step S220: Obtain the road edge data corresponding to the target position during the path learning process as target edge data, wherein the steering wheel angle at the target position is less than a preset value.

[0044] In this embodiment, the travel trajectory data includes the vehicle's steering wheel angle during the path learning process. After obtaining the road edge data and travel trajectory data during the path learning process, the vehicle can first determine the portion of the road edge data and travel trajectory data corresponding to when the vehicle is on a straight section from this data. This data is then used to subsequently determine the road boundary line and the planned straight path for the straight section based on this portion of the data. Specifically, the travel trajectory data collected by the vehicle during the path learning process includes the vehicle's steering wheel angle. Therefore, the vehicle can determine whether it is currently on a turning section based on the size of the steering wheel angle. For example, the vehicle can be pre-set to consider the vehicle as being on a straight section if the steering wheel angle is less than 45°. Thus, after determining the straight section in the vehicle's path learning process based on the steering wheel angle in the travel trajectory data, the road edge data corresponding to the straight section can be used as target edge data to subsequently determine the road boundary line corresponding to the straight section based on the target edge data.

[0045] It's worth noting that if a vehicle is on a straight road but there's an obstacle on the right edge of the road, the vehicle might turn left to bypass the obstacle during path learning. Although the vehicle is still on a straight road, based on the steering wheel angle in the trajectory data, it might be classified as being on a turning section. In this case, the vehicle should still determine the straight road section based on the steering wheel angle. However, for straight road sections with a turning trajectory, the vehicle can disregard this portion of the trajectory data and road edge data to increase the accuracy of the final straight-line planning path.

[0046] In some implementations, the vehicle can acquire the steering wheel angle in real time using sensors mounted on the steering wheel, and can also acquire the vehicle's position information using an IMU sensor. The vehicle can then combine the steering wheel angle at any given moment, the vehicle's position information, and the current time information to form a single trajectory data set. In other words, the trajectory data obtained during the vehicle's path learning process can include the vehicle's steering wheel angle, vehicle coordinates, and timestamp data. In some implementations, the trajectory data acquired by the vehicle can also include the vehicle's heading angle data, which is the angle between the vehicle's center of mass velocity direction and the X-axis of the world coordinate system. It is important to note that the vehicle's heading angle data is different from the steering wheel angle. The heading angle information represents the angle between the vehicle's overall centerline and the X-axis of the world coordinate system, while the steering wheel angle represents the rotation angle of the vehicle's steering wheel at the current moment. Rotation of the steering wheel does not necessarily mean a change in the vehicle's heading angle; the magnitudes and meanings of the two are different.

[0047] In some implementations, the road edge data and travel trajectory data acquired by the vehicle through various sensors can be coordinate data in either the vehicle coordinate system or the world coordinate system. To facilitate determining the correspondence between the road edge data and the travel trajectory data, and subsequently determining the road boundary lines and straight-line planning paths for straight sections, the vehicle can convert these coordinate data from different coordinate systems to coordinate data in the SL coordinate system. The SL coordinate system uses the road centerline as a reference, with the S-axis representing the direction of the road centerline and the L-axis representing the direction perpendicular to the road centerline. Converting the data from both the vehicle coordinate system and the world coordinate system to the SL coordinate system makes the correspondence between the road edge data and the travel trajectory data clearer.

[0048] Step S230: Based on the target edge data, determine the straight-line planning path corresponding to the straight-line segment of the vehicle in the path learning process.

[0049] In this embodiment, the target edge data determined by the vehicle is the road edge data corresponding to the target position in the road edge data collected by the vehicle during path learning. The target position refers to the position where the steering wheel angle is less than a preset value. That is, in the vehicle's collected trajectory data, the position where the steering wheel angle is less than the preset value is taken as the target position. In other words, the road edge data corresponding to the straight sections during the path learning process are sequentially used as the target edge data. Therefore, the vehicle can determine the planned straight route corresponding to the straight section based on the target edge data corresponding to the straight section. It is understandable that during data collection, even if the vehicle passes through a straight section, the collected trajectory may not be straight. The user may avoid obstacles on the road edge or yield to vehicles behind, causing the actual trajectory data to curve. In this case, the vehicle obviously cannot directly use the collected trajectory data as its planned straight route for the straight section. Therefore, the vehicle can indirectly determine the planned straight route based on the road edge data collected during path learning. Clearly, even if the vehicle's trajectory data on a straight section of road is curved, the road edge data collected on that section will not be curved. Therefore, the planned straight path derived from this road edge data will also be a straight route—the optimal route for the vehicle when there are no obstacles on that straight section. It's important to note that the planned straight path based on the target edge data is merely a reference route for the vehicle while traveling on a straight section. The vehicle will still need to avoid obstacles based on actual road conditions, but the planned straight path determined by the target edge data can still reduce unnecessary detours.

[0050] In some implementations, such as Figure 4 As shown, the method for determining the straight-ahead planning path corresponding to the straight-ahead segment of the vehicle during the path learning process based on the target edge data in step S230 can be implemented in the following way:

[0051] Step S231: Based on the coordinate data corresponding to the first edge data and the second edge data in the road coordinate system, determine the target travel point corresponding to the straight road segment of the vehicle in the path learning process. The coordinate data corresponding to the target travel point in the road coordinate system is the average value of the coordinate data corresponding to the first edge data and the second edge data.

[0052] In this embodiment, the target edge data includes first edge data and second edge data located on both sides of the road. Clearly, the road edge data acquired by the vehicle during path learning includes edge data on both sides of the road. For any target position in the vehicle's collected trajectory data where the steering wheel angle is less than a preset value, there can be two corresponding edge data points in the road edge data: the first edge data and the second edge data located on both sides of the road. The vehicle can obtain the coordinate data of the target travel point corresponding to the target position in the straight-line planning path corresponding to the straight-line segment based on the average of the coordinate data corresponding to the first edge data and the second edge data in the road coordinate system. Specifically, as... Figure 5 As shown, the black dots within the dashed boxes on the left and right sides represent road edge data collected during vehicle path learning. These include the first edge data on the left side of the road and the second edge data on the right side. The white dot in the middle represents the vehicle's trajectory data collected during path learning. It can be seen that the vehicle is currently traveling on a straight section of road, but the trajectory data is not a perfectly straight line. Therefore, the vehicle can obtain the straight-line planning path corresponding to this straight section of road using the first and second edge data. This involves taking the average of the coordinate data corresponding to the first and second edge data to obtain the coordinate data of the black dot in the middle, which represents the target travel point.

[0053] Step S232: Perform straight-line fitting on all target travel points corresponding to the same straight-line segment of the vehicle during the path learning process to obtain the straight-line planned path corresponding to the straight-line segment of the vehicle during the path learning process.

[0054] In this embodiment, the vehicle can obtain the coordinate data of the target travel point in the road coordinate system based on the first edge data and the second edge data corresponding to the target position in the target edge data. Obviously, the vehicle's travel trajectory data on a straight road segment will include multiple consecutive target positions with steering wheel angles less than a preset value. Therefore, based on the first edge data and the second edge data corresponding to each target position, the target travel point corresponding to each target position can be obtained. Due to potential errors in acquiring road edge data, and to reduce the curvature of the straight-line planning path corresponding to the straight road segment, the vehicle can perform straight-line fitting on multiple consecutive target travel points to obtain a straight line passing through the most target travel points as the straight-line planning path corresponding to that straight road segment. The vehicle can use methods such as least squares, optimization methods, or nonlinear optimization solvers to ensure that the sum of squared deviations between the final straight-line planning path and each target travel point is minimized.

[0055] Step S240: Based on the straight-line planning path and the target edge data, determine the road boundary line corresponding to the straight-line segment.

[0056] In this embodiment, using the above method, the vehicle can obtain the planned straight-ahead path corresponding to each straight-ahead segment during the path learning process. Subsequently, to determine the planned turning path connecting the various straight-ahead segments, the vehicle can also pre-determine the road boundary line of the straight-ahead segment based on the already determined planned straight-ahead path and target edge data. It is understood that the target edge data corresponding to the straight-ahead segment collected by the vehicle during the path learning process is discrete coordinate data. Therefore, based on the planned straight-ahead path corresponding to the straight-ahead segment and the discrete target edge data, the vehicle can determine a straight line parallel to the planned straight-ahead path and with the smallest sum of squared deviations from each discrete target edge data, and use this line as the road boundary line corresponding to the straight-ahead segment.

[0057] Specifically, vehicles can obtain the road boundary line corresponding to the straight-ahead section in the following way:

[0058] Based on the straight-line planning route, straight-line fitting is performed on all target edge data on the same side of the same straight-line segment to obtain the road boundary lines corresponding to both sides of the straight-line segment.

[0059] Understandably, the target edge data collected by the vehicle for the straight-ahead road segment includes the first edge data and the second edge data located on both sides of the road. When determining the road boundary line, the vehicle needs to determine the road boundary line based on the target edge data on the same side of the road. Specifically, the vehicle can perform straight-line fitting on the target edge data on the same side based on the planned straight-ahead path corresponding to the straight-ahead road segment. This yields a straight line that is parallel to the planned straight-ahead path and has the smallest sum of squared deviations from each target edge data point. This straight line is then used as the road boundary line corresponding to the straight-ahead road segment.

[0060] Step S250: Based on the road boundary line, determine the turning area where the vehicle is located in the turning segment during the path learning process.

[0061] In this embodiment, after determining the road boundary line corresponding to the straight-ahead road segment, the vehicle can determine the turning area between the two adjacent straight-ahead planned paths based on the road boundary lines corresponding to each of the two adjacent straight-ahead planned paths. It is understood that if the two straight-ahead planned paths are connected, the vehicle will inevitably pass through a turning segment during its journey from one straight-ahead planned path to the other. The vehicle can determine the turning area of ​​the turning segment between these two straight-ahead planned paths based on the road boundary lines corresponding to each of the two straight-ahead planned paths, in order to subsequently determine the vehicle's turning planned path within the turning area.

[0062] In some implementations, such as Figure 6 As shown, the step S250, which determines the turning area of ​​the turning segment in the vehicle path learning process based on the road boundary line, can also be achieved in the following way:

[0063] Step S251: Based on the intersection of the road boundary lines on the same side of the vehicle during the path learning process, determine the turning diagonal point corresponding to the turning area.

[0064] Step S252: Determine the turning area based on the diagonal turning point.

[0065] In this embodiment, the road boundary lines determined by the vehicle include the road boundary lines on both sides of the road. When determining the turning area based on the road boundary lines, the vehicle can determine the diagonal turning point of the turning area based on the road boundary lines on the same side of the road boundary lines corresponding to two adjacent straight-ahead planned paths, and then determine the turning area connecting these two straight-ahead planned paths based on the diagonal turning point. Specifically, as shown... Figure 7 As shown in the diagram, the two black dashed lines represent two adjacent straight-ahead planned paths. The parallel black solid lines on either side of these paths are the road boundary lines corresponding to the straight-ahead planned paths. The road boundary lines corresponding to the left straight-ahead planned path are the first and second boundary lines, while those corresponding to the right straight-ahead planned path are the third and fourth boundary lines. Clearly, the first and third boundary lines are on the same side, as are the second and fourth boundary lines. Therefore, a vehicle can determine one diagonal turning point of the turning area based on the first and third boundary lines, and the other diagonal turning point based on the second and fourth boundary lines. This allows us to obtain... Figure 7 The vehicle can then use diagonal points 1 and 2 to determine its turning area between the two planned straight paths. This means, based on diagonal points 1 and 2, the vehicle can obtain... Figure 7 The bold black rectangular area serves as the turning area.

[0066] Step S260: Based on the turning area and the straight-line planning path adjacent to the turning area, determine the turning start and end points and the turning reference point of the vehicle in the turning section.

[0067] In this embodiment, the turning start and end points include the starting point where the vehicle enters the turning area based on the straight-line planned path, and the ending point where it leaves the turning area. Normally, if the vehicle is manually driven, determining the turning path in the turning area typically only requires determining the turning start and end points—that is, the starting point for entering the turning area and the ending point for leaving the turning area. However, if the vehicle is in autonomous driving mode, due to limitations on the steering wheel's rotation angle, even if the turning start and end points are determined, the vehicle may still be unable to follow the predetermined path due to the steering wheel angle limitation. Therefore, to avoid situations where the vehicle cannot follow the planned path due to excessive steering wheel rotation during turning, the vehicle can also determine a turning reference point in the turning area. Based on this turning reference point and the turning start and end points, the turning trajectory can be determined, ensuring that the steering wheel rotation angle is always less than a preset value during autonomous driving while following the planned turning path.

[0068] In some implementations, such as Figure 8 As shown, in step S260, based on the turning area and the straight-line planning path adjacent to the turning area, the starting and ending points of the turn and the turning reference point corresponding to the vehicle in the turning area can also be achieved in the following way:

[0069] Step S261: Obtain the straight-line planning path adjacent to the turning area as the target planning path.

[0070] Step S262: Based on the intersection point of the target planned path and the turning area, determine the starting and ending points of the turn corresponding to the vehicle in the turning section.

[0071] In this embodiment, the turning area determined by the vehicle is necessarily the area connecting two different straight-ahead planned paths. Therefore, the vehicle can use the two straight-ahead planned paths adjacent to the turning area as target planned paths, and then determine the turning planned path corresponding to the vehicle on the turning segment based on the two target planned paths. Specifically, the vehicle can determine the intersection points of the two target planned paths with the turning area, and use these two intersection points as the start and end points of the vehicle on the turning segment, in order to determine the turning planned path based on the start and end points of the turning segment. For example, please refer again to... Figure 7 Vehicles can use the intersection of the two planned straight paths with the turning area, which is the intersection of the two white triangles in the figure, as the starting and ending points of the turn on the turning section.

[0072] Step S263: Based on the intersection of the two intersecting straight-line planned paths, determine the turning reference point of the vehicle on the turning section.

[0073] In this embodiment, after determining the start and end points of the turning segment based on the straight-line planning path adjacent to the turning area, the vehicle can also determine a turning reference point based on the straight-line planning path. This turning reference point is used to ensure that the vehicle can smoothly pass through the turn start and end points in autonomous driving mode, and that the steering wheel angle never exceeds a preset value during the driving process. It is understood that if the vehicle's steering wheel angle is not restricted, i.e., the vehicle is not in autonomous driving mode, the steering wheel angle in the vehicle's trajectory passing through the turn start and end points may exceed the preset value. In other words, if the vehicle's steering wheel angle is restricted, the vehicle may not be able to pass through the turn start and end points of the turning segment simultaneously using the same trajectory. Therefore, after determining the turn start and end points of the turning area, the vehicle can also determine a turning reference point within the turning area based on the two straight-line planning paths adjacent to the turning area. The turning planning trajectory determined based on the turn start and end points of the turning area and the turning reference point ensures that the steering wheel angle never exceeds the preset value during the path learning process based on the turning planning trajectory, thus making it applicable to autonomous driving. Specifically, the vehicle can use the intersection of two planned straight-line paths adjacent to the turning area as a turning reference point. Clearly, this turning reference point is located within the turning area. Please refer to this again. Figure 7 The intersection point shown by the white pentagon is the intersection point of the two straight-line planned paths adjacent to the turning area. Vehicles can use this intersection point as the turning reference point for the turning section.

[0074] Step S270: Based on the starting and ending points of the turn corresponding to the turn area and the turn reference point, determine the turn planning path of the vehicle corresponding to the turn segment.

[0075] In this embodiment, after determining the starting and ending points and the reference point corresponding to the turning area through the above steps, the vehicle can determine the planned turning path for the turning segment based on these points. The planned turning path should pass through both the starting and ending points and the reference point. This determined turning path ensures that the vehicle can smoothly pass through the turning segment under autonomous driving conditions, and that the steering wheel rotation angle does not exceed a preset value. For example, please refer to... Figure 9This diagram illustrates the planned turning path for a vehicle on a turning section. The black circles indicate the intersections of two adjacent straight-line planned paths and the intersections of these two straight-line planned paths with the turning area; these are the turning reference points and the start and end points of the turn. It can be seen that if the steering wheel angle is unrestricted (i.e., the vehicle is manually driven), the vehicle's trajectory through the turning section can be path 1, which only passes through the start and end points of the turning area. However, if the steering wheel angle is restricted (i.e., the vehicle is in autonomous driving mode), the vehicle may not be able to follow path 1 because the steering wheel angle may exceed a preset value in some areas. In this case, to ensure the vehicle's actual trajectory passes through the start and end points of the turning area, the vehicle can also determine the turning reference point based on the two straight-line planned paths and ensure that the planned turning path passes through the turning reference point, i.e., determine... Figure 9 If the vehicle travels along path 2, the steering wheel angle will always be less than the preset value during the journey.

[0076] Step S280: Based on the straight-line planning path and the turning planning path, determine the driving path of the vehicle during the path learning process.

[0077] In this embodiment, step S280 can be referred to the content of other embodiments, and will not be repeated here.

[0078] The path planning method provided in this application acquires road edge data and travel trajectory data collected during vehicle path learning. Based on the road edge data corresponding to the straight-ahead road segment during the vehicle path learning process, it determines the straight-ahead planned path for the vehicle on the straight-ahead segment. Then, based on the straight-ahead planned path and target edge data, it determines the road boundary line corresponding to the straight-ahead segment. Furthermore, based on the road boundary line, it determines the turning area corresponding to the turning segment during the path learning process. Based on the straight-ahead planned path adjacent to the turning area, it determines the turning planned path for the vehicle on the turning segment. Finally, by connecting the straight-ahead planned path and the turning planned path, it forms the vehicle's driving path during the path learning process. By pre-acquiring the road edge data and travel trajectory data during the vehicle path learning process through manual driving, and determining the corresponding travel path for the vehicle traveling on the same road segment during autonomous driving, it can reduce unnecessary detours in the vehicle's travel path, making the vehicle's travel path more consistent with actual road conditions.

[0079] Please see Figure 10The diagram illustrates a structural block diagram of a path planning device 200 provided in one embodiment of this application. The path planning device 200 includes: a data acquisition module 210, a straight path module 220, a turning path module 230, and a driving path module 240. Specifically, the data acquisition module 210 is used to acquire road edge data and travel trajectory data collected during the path learning process; the straight path module 220 is used to determine the road boundary line corresponding to the straight section during the path learning process and the straight-line planned path for the vehicle on the straight section based on the road edge data and travel trajectory data; the turning path module 230 is used to determine the turning planned path corresponding to the turning section during the path learning process based on the road boundary line and the straight-line planned path; and the driving path module 240 is used to determine the driving path for the vehicle during the driving process based on the straight-line planned path and the turning planned path.

[0080] In one possible implementation, the travel trajectory data includes the vehicle's steering wheel angle during path learning. The straight-path module 220 includes an edge data acquisition unit, a straight-path planning unit, and a boundary line determination unit. The edge data acquisition unit acquires road edge data at the target location during path learning as target edge data, wherein the steering wheel angle at the target location is less than a preset value. The straight-path planning unit determines the straight-path planning route corresponding to the straight-path segment during path learning based on the target edge data. The boundary line determination unit determines the road boundary line corresponding to the straight-path segment based on the straight-path planning route and the target edge data.

[0081] As one possible implementation, the target edge data includes first edge data and second edge data located on both sides of the road, respectively. The straight path planning unit is also used to determine the target travel point corresponding to the straight road segment of the vehicle during the path learning process based on the coordinate data corresponding to the first edge data and the second edge data in the road coordinate system. The coordinate data corresponding to the target travel point in the road coordinate system is the average value of the coordinate data corresponding to the first edge data and the second edge data. Straight line fitting is performed on all target travel points corresponding to the same straight road segment of the vehicle during the path learning process to obtain the straight planning path corresponding to the straight road segment of the vehicle during the path learning process.

[0082] As one possible implementation, the boundary line determination unit is also used to perform straight line fitting on all target edge data on the same side of the same straight road segment based on the straight planning route, so as to obtain the road boundary lines corresponding to both sides of the straight road segment.

[0083] In one possible implementation, the turning path module 230 includes a turning area determination unit, an intermediate point determination unit, and a turning path determination unit. The turning area determination unit determines the turning area where the vehicle is located during the path learning process, based on road boundary lines. The intermediate point determination unit determines the turning start and end points and a turning reference point for the vehicle on the turning segment, based on the turning area and the adjacent straight-line planned path. The turning start and end points include the starting point where the vehicle enters the turning area based on the straight-line planned path and the ending point where it leaves the turning area. The turning path determination unit determines the turning planned path for the vehicle on the turning segment, based on the turning start and end points and the turning reference point corresponding to the turning area.

[0084] As one possible implementation, the turning area determination unit is also used to determine the turning diagonal point corresponding to the turning area based on the intersection of the road boundary lines on the same side of the vehicle during the path learning process; and to determine the turning area based on the turning diagonal point.

[0085] As one possible implementation, a straight-line planning path adjacent to the turning area is obtained as the target planning path; based on the intersection point of the target planning path and the turning area, the starting and ending points of the turn for the vehicle in the turning segment are determined; based on the intersection point of the two intersecting straight-line planning paths, the turning reference point for the vehicle in the turning segment is determined.

[0086] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described device and module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0087] In the several embodiments provided in this application, the coupling between modules can be electrical, mechanical, or other forms of coupling.

[0088] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0089] In summary, the solution provided in this application acquires road edge data and travel trajectory data during the vehicle's path learning process. Based on the road edge data and travel trajectory data, it determines the road boundary line corresponding to the straight-ahead segment and the vehicle's planned straight-ahead path for that segment. Based on the road boundary line and the planned straight-ahead path, it determines the planned turning path for the vehicle's turning segment. Based on the planned straight-ahead path and the planned turning path, it determines the vehicle's driving path during the entire journey. By using the data collected during the vehicle's path learning process, it determines the planned straight-ahead and planned turning paths for the straight-ahead and turning segments, respectively, thereby obtaining the vehicle's driving path and making it more consistent with the actual road environment.

[0090] Please refer to Figure 11 The diagram illustrates a structural block diagram of a vehicle 400 provided in an embodiment of this application. The vehicle 400 in this application may include one or more of the following components: a processor 410, a memory 420, and one or more application programs. The one or more application programs may be stored in the memory 420 and configured to be executed by one or more processors 410. The one or more programs are configured to perform the methods described in the foregoing method embodiments.

[0091] Processor 410 may include one or more processing cores. Processor 410 connects to various parts of the computer device using various interfaces and lines, and performs various functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in memory 420, and by calling data stored in memory 420. Optionally, processor 410 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). Processor 410 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the displayed content; and the modem handles wireless communication. It is understood that the modem may also not be integrated into processor 410 and may be implemented separately using a communication chip.

[0092] The memory 420 may include random access memory (RAM) or read-only memory (ROM). The memory 420 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 420 may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), and instructions for implementing the various method embodiments described below. The data storage area may also store data created during the use of the computer device (such as phone books, audio and video data, chat log data, etc.).

[0093] Please refer to Figure 12 This diagram illustrates a structural block diagram of a computer-readable storage medium provided in an embodiment of this application. The computer-readable medium 800 stores program code that can be called by a processor to execute the methods described in the above method embodiments.

[0094] The computer-readable storage medium 800 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Optionally, the computer-readable storage medium 800 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 800 has storage space for program code 810 that performs any of the method steps described above. This program code can be read from or written to one or more computer program products. The program code 810 may be compressed, for example, in a suitable form.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A path planning method, characterized in that, The method includes: When entering a specific road segment, the vehicle acquires road edge data and travel trajectory data collected during the path learning process. The path learning process is the path learning process when the vehicle passes through the specific road segment under manual driving conditions. Based on the road edge data and travel trajectory data, the road boundary line corresponding to the straight road segment in the path learning process and the straight planning path of the vehicle corresponding to the straight road segment are determined. Based on the road boundary line and the straight-line planning path, the turning planning path corresponding to the turning segment of the vehicle in the path learning process is determined. Based on the straight-line planning path and the turning planning path, the driving path of the vehicle during the path learning process is determined.

2. The method according to claim 1, characterized in that, The travel trajectory data includes the steering wheel angle of the vehicle during the path learning process. The process of determining the road boundary line corresponding to the straight section during the path learning process and the planned straight path for the vehicle on the straight section, based on the road edge data and the travel trajectory data, includes: The road edge data corresponding to the target position during the path learning process is obtained as target edge data, wherein the steering wheel angle at the target position is less than a preset value; Based on the target edge data, the straight-line planning path corresponding to the straight-line segment of the vehicle in the path learning process is determined; Based on the straight-line planning path and the target edge data, the road boundary line corresponding to the straight-line segment is determined.

3. The method according to claim 2, characterized in that, The target edge data includes first edge data and second edge data located on both sides of the road, respectively. The step of determining the straight-ahead planned path for the vehicle corresponding to the straight-ahead segment during the path learning process based on the target edge data includes: Based on the coordinate data corresponding to the first edge data and the second edge data in the road coordinate system, the target travel point corresponding to the straight road segment of the vehicle in the path learning process is determined, and the coordinate data corresponding to the target travel point in the road coordinate system is the average value of the coordinate data corresponding to the first edge data and the second edge data. For all target travel points corresponding to the same straight road segment during the path learning process, a straight-line fitting is performed to obtain the straight-line planned path corresponding to the straight road segment during the path learning process.

4. The method according to claim 2, characterized in that, The step of determining the road boundary line corresponding to the straight road segment based on the straight-ahead planned path and the target edge data includes: Based on the straight-line planning path, straight-line fitting is performed on all target edge data on the same side of the same straight-line segment to obtain the road boundary lines corresponding to both sides of the straight-line segment.

5. The method according to any one of claims 1-4, characterized in that, The step of determining the turning planning path corresponding to the turning segment of the vehicle in the path learning process, based on the road boundary line and the straight-ahead planning path, includes: Based on the road boundary line, the turning area where the vehicle is located during the turning segment in the path learning process is determined; Based on the turning area and the straight-line planning path adjacent to the turning area, the turning start and end points and the turning reference point corresponding to the vehicle in the turning section are determined. The turning start and end points include the turning start point when the vehicle enters the turning area based on the straight-line planning path, and the turning end point when it leaves the turning area. Based on the starting and ending points of the turn and the turning reference point corresponding to the turning area, the turning planning path of the vehicle in the turning section is determined.

6. The method according to claim 5, characterized in that, The step of determining the turning area of ​​the vehicle in the turning segment during the path learning process based on the road boundary line includes: Based on the intersection of the road boundary lines on the same side of the vehicle during the path learning process, the turning diagonal point corresponding to the turning area is determined. The turning area is determined based on the diagonal turning point.

7. The method according to claim 5, characterized in that, The process of determining the starting and ending points of the turn and the turning reference point for the vehicle on the turning segment based on the turning area and the straight-line planning path adjacent to the turning area includes: Obtain the straight-line planning path adjacent to the turning area as the target planning path; Based on the intersection point of the target planned path and the turning area, determine the starting and ending points of the turn for the vehicle on the turning section; Based on the intersection of the two intersecting straight-line planned paths, the turning reference point corresponding to the vehicle in the turning segment is determined.

8. A path planning device, characterized in that, The device includes: The data acquisition module is used to acquire road edge data and travel trajectory data collected during the path learning process when the vehicle enters a specific road segment. The path learning process is the path learning process when the vehicle passes through the specific road segment under manual driving conditions. The straight path module is used to determine the road boundary line corresponding to the straight section in the path learning process and the straight planned path of the vehicle corresponding to the straight section based on the road edge data and the travel trajectory data. A turning path module is used to determine the turning planning path corresponding to the turning segment of the vehicle in the path learning process, based on the road boundary line and the straight-going planned path. The driving path module is used to determine the driving path of the vehicle during the path learning process based on the straight-line planned path and the turning planned path.

9. A vehicle, characterized in that, include: One or more processors; Memory; One or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs being configured to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code that can be invoked by a processor to execute the method as described in any one of claims 1-7.

Images