Travel path planning method and device, vehicle, storage medium and terminal

Abstract

The application discloses a driving path planning method and device, a vehicle, a storage medium and a terminal, relates to the technical field of automatic driving, and mainly aims to solve the problem of low effectiveness of the existing driving path planning. The method comprises the following steps: determining the path boundary of the vehicle, and performing path planning decision processing of the vehicle according to the path boundary; the path boundary is generated according to the collision avoidance distance of the trajectory normal of the vehicle and the overlapping angle point position of the vehicle and the obstacle expected to collide; under the condition that the path planning decision processing fails according to the path boundary, the path boundary is adjusted and constrained, and the path planning decision processing is performed based on the path boundary after the constraint adjustment; under the condition that the path planning decision processing fails based on the path boundary after the constraint adjustment, the reference driving trajectory of the vehicle is acquired, and the driving path of the path planning decision is determined according to the reference driving trajectory.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, and in particular, to a driving path planning method and device, a vehicle, a storage medium, and a terminal. BACKGROUND

[0002] With the rapid development of vehicle automatic driving technology, the control of vehicles is becoming more and more refined. In order to realize the automatic control of vehicles, the driving path of the vehicle is usually planned for automatic driving. In the process of driving path planning for large-curvature turning, a constraint condition is usually selected as a path boundary, so that the driving path obtained through path planning decision avoids collision with obstacles.

[0003] At present, in the process of lateral path planning for large-curvature turning, the path boundary is usually set by the detected avoidance space, so as to solve the path planning decision. However, due to problems such as coordinate projection and large-curvature optimization in the solving process, the solving often fails, and the driving path after path planning decision cannot be accurately obtained, so that the vehicle cannot perform collision avoidance control based on the result of path planning decision, greatly reducing the effectiveness of driving path planning. SUMMARY

[0004] Therefore, the present application provides a driving path planning method and device, a vehicle, a storage medium, and a terminal, which mainly aims to solve the problem of low effectiveness of existing driving path planning.

[0005] According to one aspect of the present application, a driving path planning method is provided, comprising:

[0006] determining a path boundary of a vehicle, and performing path planning decision processing of the vehicle according to the path boundary, wherein the path boundary is generated according to an avoidance distance of an expected collision between the vehicle and an obstacle, which is determined according to a trajectory normal of the vehicle and an overlapping angle point position;

[0007] under the condition that the path planning decision processing according to the path boundary fails, adjusting the constraint of the path boundary, and performing path planning decision processing based on the constraint-adjusted path boundary;

[0008] under the condition that the path planning decision processing based on the constraint-adjusted path boundary fails, obtaining a reference driving trajectory of the vehicle, and determining a driving path of path planning decision according to the reference driving trajectory.

[0009] Further, the adjustment of the constraint of the path boundary and the path planning decision processing based on the constraint-adjusted path boundary comprise:

[0010] configure a loss term weight corresponding to the obstacle in the path boundary as a first constraint adjustment parameter, and perform path planning decision processing based on the path boundary containing the first constraint adjustment parameter;

[0011] In the case of failure of path planning decision processing according to the path boundary containing the first constraint adjustment parameter, configure loss term weights corresponding to all constraint objects in the path boundary as a second constraint adjustment parameter, and perform path planning decision processing based on the path boundary containing the second constraint adjustment parameter.

[0012] Further, after the path planning decision processing based on the path boundary containing the first constraint adjustment parameter, the method further comprises:

[0013] In the case of success of path planning decision processing according to the path boundary containing the first constraint adjustment parameter, determine the driving path of path planning decision processing.

[0014] Further, the method further comprises:

[0015] In the case of detecting that the vehicle is expected to collide with the obstacle in the driving path, re-determine the path boundary of the vehicle to perform path planning decision processing of the vehicle according to the re-determined path boundary.

[0016] Further, the determination of the path boundary of the vehicle comprises:

[0017] In the first coordinate system, determine the overlapping corner point position expected to collide according to the profile point positions corresponding to the vehicle and the obstacle in the driving path, respectively;

[0018] Determine the profile corner point overlapping line based on the trajectory normal of the vehicle, the overlapping corner point position, and the decision corner point reference line, the decision corner point reference line being parallel to the trajectory normal, and the profile corner point overlapping line overlapping and covering the profile edge of the vehicle or the obstacle expected to collide;

[0019] If the first intersection position between the profile corner point overlapping line and the decision corner point reference line is not on the profile edge, determine the collision avoidance distance of the vehicle based on the overlapping corner point position, the decision corner point reference line, and the profile corner point position on the profile corner point overlapping line;

[0020] In the second coordinate system, generate a path boundary for path planning decision based on the collision avoidance distance.

[0021] Further, the overlap corner point position comprises at least one main overlap corner point position and two auxiliary overlap corner point positions, and the determining the collision avoidance distance of the vehicle based on the overlap corner point position, the decision corner reference line and the contour corner point position on the contour corner overlap line comprises:

[0022] If the main overlap corner point position is the same as the vehicle contour corner point position of the vehicle, a collision avoidance translation line is constructed based on the main overlap corner point position, the obstacle contour corner point position and the auxiliary overlap corner point position, and a second intersection position between the main overlap corner point position, the collision avoidance translation line and the decision corner reference line is determined as the collision avoidance distance of the vehicle, and the contour corner overlap line overlaps the contour edge of the vehicle on which the collision is expected to occur and the vehicle contour corner point position.

[0023] If the main overlap corner point position is the same as the obstacle contour corner point position of the obstacle, a collision avoidance translation line is constructed based on the main overlap corner point position, the vehicle contour corner point position and the auxiliary overlap corner point position, and a second intersection position between the main overlap corner point position, the collision avoidance translation line and the decision corner reference line is determined as the collision avoidance distance of the vehicle, and the contour corner overlap line overlaps the contour edge of the vehicle on which the collision is expected to occur and the vehicle contour corner point position.

[0024] Further, the collision avoidance translation line is constructed based on the main overlap corner point position, the vehicle contour corner point position or the obstacle contour corner point position and the auxiliary overlap corner point position, and a second intersection position between the main overlap corner point position, the collision avoidance translation line and the decision corner reference line is determined as the collision avoidance distance of the vehicle, and the collision avoidance translation line is constructed based on the slope of the reference translation line and the vehicle contour corner point position or the obstacle contour corner point position.

[0025] The target auxiliary overlap corner point position is selected based on the vector direction of the auxiliary overlap corner point position and the main overlap corner point position relative to the trajectory normal line.

[0026] The reference translation line is constructed based on the target auxiliary overlap corner point position and the main overlap corner point position, and the collision avoidance translation line is constructed based on the slope of the reference translation line and the vehicle contour corner point position or the obstacle contour corner point position.

[0027] The second intersection position between the collision avoidance translation line and the decision corner reference line is determined, and the distance between the second intersection position and the main overlap corner point position is determined as the collision avoidance distance.

[0028] Further, the collision avoidance translation line is constructed based on the slope of the reference translation line and the vehicle contour corner point position or the obstacle contour corner point position, and the collision avoidance translation line is constructed based on the slope of the reference translation line and the vehicle contour corner point position or the obstacle contour corner point position.

[0029] According to a vector direction of the vehicle profile corner point position or the obstacle profile corner point position and the main overlap corner point position relative to the trajectory normal line, a target vehicle profile corner point position or a target obstacle profile corner point position is selected, and an avoidance translation line is constructed according to a slope of the reference translation line and the target vehicle profile corner point position or the target obstacle profile corner point position.

[0030] Further, the determination of the profile corner point overlap line based on the trajectory normal line of the vehicle and the overlap corner point position, and the decision corner point reference line includes:

[0031] The profile corner point overlap line is constructed according to two auxiliary overlap corner point positions in the overlap corner point position, and the decision corner point reference line is constructed according to a slope of the trajectory normal line based on a main overlap corner point position in the overlap corner point position.

[0032] According to another aspect of the present application, a driving path planning device is provided, which includes:

[0033] A determination module is configured to determine a path boundary of a vehicle, and to perform path planning decision processing of the vehicle according to the path boundary, the path boundary being generated according to an avoidance distance determined based on a trajectory normal line of the vehicle and an overlap corner point position at which the vehicle is expected to collide with an obstacle;

[0034] An adjustment module is configured to, in a condition that the path planning decision processing fails according to the path boundary, perform constraint adjustment on the path boundary, and to perform path planning decision processing based on the path boundary after the constraint adjustment;

[0035] An acquisition module is configured to, in a condition that the path planning decision processing fails based on the path boundary after the constraint adjustment, acquire a reference driving trajectory of the vehicle, and to determine a driving path of the path planning decision according to the reference driving trajectory.

[0036] Further, the adjustment module includes:

[0037] A first configuration unit is configured to configure a loss term weight corresponding to the obstacle in the path boundary as a first constraint adjustment parameter, and to perform path planning decision processing based on the path boundary containing the first constraint adjustment parameter;

[0038] A second configuration unit is configured to, in a condition that the path planning decision processing fails according to the path boundary containing the first constraint adjustment parameter, configure loss term weights corresponding to all constraint objects in the path boundary as a second constraint adjustment parameter, and to perform path planning decision processing based on the path boundary containing the second constraint adjustment parameter.

[0039] Further, the determining module is further configured to determine the driving path of the path planning decision processing under a condition that the path planning decision processing according to the path boundary containing the first constraint adjustment parameter is successful.

[0040] Further, the determining module is further configured to re-determine the path boundary of the vehicle under a condition that the vehicle is expected to collide with the obstacle in the driving path, so as to perform the path planning decision processing of the vehicle according to the re-determined path boundary.

[0041] Further, the determining module comprises:

[0042] A first determining unit is configured to determine, in a first coordinate system, an overlapping corner point position at which the vehicle is expected to collide with the obstacle according to contour point positions of the vehicle and the obstacle in the driving path respectively.

[0043] A second determining unit is configured to determine a contour corner point overlapping line based on a trajectory normal of the vehicle, the overlapping corner point position, and a decision corner point reference line, the decision corner point reference line being parallel to the trajectory normal, and the contour corner point overlapping line overlapping and covering a contour edge at which the vehicle or the obstacle is expected to collide.

[0044] A third determining unit is configured to determine an avoidance distance of the vehicle based on the overlapping corner point position, the decision corner point reference line, and a contour corner point position on the contour corner point overlapping line, if a first intersection position between the contour corner point overlapping line and the decision corner point reference line is not on the contour edge.

[0045] A generating unit is configured to generate, in a second coordinate system, a path boundary for path planning decision based on the avoidance distance.

[0046] Further, the third determining unit is specifically configured to: if the main overlap corner point position is the same as the vehicle contour corner point position of the vehicle, construct an avoidance translation line based on the main overlap corner point position, the obstacle contour corner point position, and the auxiliary overlap corner point position, and determine the avoidance distance of the vehicle according to a second intersection position between the main overlap corner point position and the avoidance translation line and the decision corner reference line, the contour corner point overlap line being overlapped on the contour edge of the obstacle on which the obstacle is expected to collide and the obstacle contour corner point position; and if the main overlap corner point position is the same as the obstacle contour corner point position of the obstacle, construct an avoidance translation line based on the main overlap corner point position, the vehicle contour corner point position, and the auxiliary overlap corner point position, and determine the avoidance distance of the vehicle according to a second intersection position between the main overlap corner point position and the avoidance translation line and the decision corner reference line, the contour corner point overlap line being overlapped on the contour edge of the vehicle on which the vehicle is expected to collide and the vehicle contour corner point position.

[0047] Further, the third determining unit is specifically configured to: select a target auxiliary overlap corner point position based on the vector direction of the auxiliary overlap corner point position and the main overlap corner point position relative to the trajectory normal line; construct a reference translation line according to the target auxiliary overlap corner point position and the main overlap corner point position, and construct an avoidance translation line based on the slope of the reference translation line and the vehicle contour corner point position or the obstacle contour corner point position; determine a second intersection position between the avoidance translation line and the decision corner reference line, and determine the avoidance distance as the distance between the second intersection position and the main overlap corner point position.

[0048] Further, the third determining unit is specifically configured to: select a target vehicle contour corner point position or a target obstacle contour corner point position based on the vector direction of the vehicle contour corner point position or the obstacle contour corner point position relative to the main overlap corner point position, and construct an avoidance translation line according to the slope of the reference translation line and the target vehicle contour corner point position or the target obstacle contour corner point position.

[0049] Further, the second determining unit is specifically configured to: construct a contour corner point overlap line according to the two auxiliary overlap corner point positions in the overlap corner point positions, and construct a decision corner reference line according to the slope of the trajectory normal line based on the main overlap corner point position in the overlap corner point positions.

[0050] According to another aspect of the present application, a storage medium is provided, and the storage medium stores at least one executable instruction, and the executable instruction causes a processor to perform operations corresponding to the above-mentioned driving path planning method.

[0051] According to another aspect of the present application, a terminal is provided, comprising a processor, a memory, a communication interface and a communication bus, the processor, the memory and the communication interface accomplish communication with each other through the communication bus;

[0052] The memory is configured to store at least one executable instruction, and the executable instruction causes the processor to perform operations corresponding to the driving path planning method.

[0053] By means of the above technical solution, the technical solution provided by the embodiments of the present application has at least the following advantages:

[0054] The present application provides a driving path planning method and device, vehicle, storage medium and terminal. Compared with the prior art, the embodiments of the present application determine the path boundary of the vehicle, and make path planning decision processing of the vehicle according to the path boundary. The path boundary is generated according to the collision avoidance distance determined according to the trajectory normal of the vehicle and the overlapping angle point position of the vehicle and the obstacle. In the case that the path planning decision processing based on the path boundary fails, the path boundary is adjusted and constrained, and the path planning decision processing is made based on the adjusted and constrained path boundary. In the case that the path planning decision processing based on the adjusted and constrained path boundary fails, the reference driving trajectory of the vehicle is obtained, and the driving path of the path planning decision is determined according to the reference driving trajectory. The optimized solution result is obtained in the manner of reducing the constraint limitation, the solving effectiveness of the path planning decision is improved, and the effectiveness of the driving path planning is improved.

[0055] The above description is only a summary of the technical solutions of the present application. In order to enable one skilled in the art to better understand the technical means of the present application, the content of the specification can be implemented, and in order to enable the above and other purposes, features and advantages of the present application to be more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0056] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not meant to limit the present application. Furthermore, the same reference numerals are used throughout the several drawings to designate the same or similar parts. In the drawings:

[0057] Figure 1 A driving path planning method flowchart provided by the embodiments of the present application is shown;

[0058] Figure 2 Another driving path planning method flowchart provided by the embodiments of the present application is shown;

[0059] Figure 3A flow chart of another driving path planning method provided by the embodiment of the present application is shown;

[0060] Figure 4 A schematic diagram of a collision overlap coverage area provided by the embodiment of the present application is shown;

[0061] Figure 5 A schematic diagram of a contour angle point overlapping line covering the contour edge of an obstacle provided by the embodiment of the present application is shown;

[0062] Figure 6 A schematic diagram of a contour angle point overlapping line covering the contour edge of a vehicle provided by the embodiment of the present application is shown;

[0063] Figure 7 A schematic diagram of an auxiliary overlap angle point position and a contour angle point position provided by the embodiment of the present application is shown;

[0064] Figure 8 A schematic diagram of a collision avoidance distance determination provided by the embodiment of the present application is shown;

[0065] Figure 9 Another schematic diagram of a collision avoidance distance determination provided by the embodiment of the present application is shown;

[0066] Figure 10 A block diagram of a driving path planning device provided by the embodiment of the present application is shown;

[0067] Figure 11 A schematic diagram of the structure of a terminal provided by the embodiment of the present application is shown. DETAILED DESCRIPTION

[0068] Exemplary embodiments of the present disclosure will be described more fully hereinafter with reference to the accompanying drawings; however, they are not limited to the embodiments set forth herein but can be implemented in various forms. The present embodiments are provided so that this disclosure will be thorough, and will fully convey the scope of the disclosure to those skilled in the art.

[0069] For lateral path planning in a large-curvature turning process, the path boundary is usually set by the detected avoidance space, so as to solve the path planning decision. However, due to problems such as coordinate projection and large-curvature optimization in the solving process, the solving often fails, and the driving path after the path planning decision cannot be accurately obtained, so that the vehicle cannot perform collision avoidance control based on the result of the path planning decision, greatly reducing the accuracy and effectiveness of the driving path planning. The embodiment of the present application provides a driving path planning method, as shown in Figure 1 The method comprises the following steps.

[0070] 101、determining a path boundary of the vehicle, and making path planning decision processing of the vehicle according to the path boundary.

[0071] In the embodiment of the present application, the vehicle will make real-time path planning decision during automatic driving to ensure that the vehicle drives automatically according to the path trajectory. At this time, the path boundary is taken as a constraint condition in the path planning decision processing to be optimized and solved, so as to obtain the driving path after optimization decision. Since the purpose of path planning decision is to avoid collision, the collision avoidance distance generated according to the trajectory normal of the vehicle and the overlapping angle point position of the vehicle and the obstacle expected to collide is determined when the path boundary of the vehicle is determined. Specifically, since the vehicle drives automatically according to the path trajectory, the trajectory normal of the vehicle driving on the path trajectory, i.e. the straight line perpendicular to the tangent direction of the path trajectory, can be determined. Moreover, the overlapping angle point position of the vehicle and the obstacle expected to collide can be determined based on the overlapping part of the profile point position of the vehicle and the profile point position of the obstacle, so as to determine the overlapping angle point position when the collision is expected to occur. At this time, the path planning decision in the embodiment of the present application can be optimized and solved by any decision optimization algorithm based on the path boundary as a constraint condition in the field of automatic driving, so as to obtain the driving path, which is not limited in the embodiment of the present application. In addition, since the driving path is the path expected to be driven by the vehicle during automatic driving, at this time, the driving path can be determined based on the pre-configured path planning content, and the path planning decision optimization solving is performed based on the path boundary as a constraint condition to obtain the driving path for collision avoidance when the collision is expected to occur. In a specific implementation scenario, since the path trajectory in path planning is pre-configured, for example, in the underground garage scenario, the path trajectory is the driving route in the underground garage, and the path planning can be performed by a lateral optimization strategy to achieve collision avoidance.

[0072] It should be noted that the vehicle is a vehicle with an automatic control system in an automatic driving scenario, including passenger cars and commercial vehicles. Common models of passenger cars include but are not limited to sedans, sport utility vehicles, multi-person business cars, etc. Common models of commercial vehicles include but are not limited to pickup trucks, microvans, self-loading vehicles, cargo trucks, tractors, trailers and mining vehicles, etc. At this time, the vehicle can realize automatic driving based on the automatic control system.

[0073] 102、in the case of failure of path planning decision processing according to the path boundary, adjusting the constraint of the path boundary, and making path planning decision processing based on the path boundary after constraint adjustment.

[0074] In the embodiment of the present application, since the path planning decision processing is optimized and solved based on the road boundary as a constraint condition, when the path planning decision processing fails with the path boundary as a constraint condition, that is, a normal solution cannot be obtained, the driving path after optimization cannot be obtained. Therefore, in order to obtain a driving path so that the vehicle can be normally controlled to automatically drive, the path boundary is adjusted as a constraint, and the path planning decision processing is re-performed based on the path boundary after the constraint adjustment.

[0075] It should be noted that in the embodiment of the present application, the path boundary is adjusted as a constraint, that is, the path boundary as a constraint condition is adjusted to relax the constraint, thereby reducing the constraint limitation in the path planning decision solving process, so that the path planning decision can obtain a normal solution.

[0076] 103、In the condition that the path planning decision processing based on the path boundary after the constraint adjustment fails, the reference driving trajectory of the vehicle is obtained, and the driving path of the path planning decision is determined according to the reference driving trajectory.

[0077] In the embodiment of the present application, when the path planning decision processing based on the path boundary after the constraint adjustment fails, that is, the path boundary after the constraint adjustment as a constraint condition cannot still obtain a normal solution. Therefore, in order to obtain an effective driving path for automatically controlling the vehicle to drive, in the embodiment of the present application, the reference driving trajectory of the vehicle is obtained, and the final driving path of the path planning decision is determined based on the reference driving trajectory.

[0078] It should be noted that in the embodiment of the present application, the reference driving trajectory can be a model trajectory of a vehicle driving model constructed in advance for the same road section in a large-curvature turning or reversing application scenario, or a historical trajectory of the vehicle normally driving on the same road section, which is not limited in the embodiment of the present application. Therefore, when the driving path is determined based on the reference driving trajectory, the reference driving trajectory can be directly used as the driving trajectory after the current time to control the vehicle to automatically drive according to the driving trajectory.

[0079] In another embodiment of the present application, in order to further limit and illustrate, as shown in Figure 2 The step of adjusting the path boundary as a constraint and performing path planning decision processing based on the path boundary after the constraint adjustment includes:

[0080] 201、The loss term weight corresponding to the obstacle in the path boundary is configured as a first constraint adjustment parameter, and path planning decision processing is performed based on the path boundary containing the first constraint adjustment parameter;

[0081] 202、in the case of failure of the path planning decision processing based on the path boundary containing the first constraint adjustment parameter, the loss term weight corresponding to all constraint objects in the path boundary is configured as a second constraint adjustment parameter, and the path planning decision processing is performed based on the path boundary containing the second constraint adjustment parameter.

[0082] In order to accurately and effectively adjust the constraint of the path boundary, so as to realize the normal solution of the path planning decision, at this time, since the path boundary as a constraint condition contains the loss term of the obstacle and the lateral distance which can be a constraint object, when adjusting the constraint of the path boundary, specifically, the loss term weight corresponding to the obstacle is configured first, so as to complete the first constraint adjustment. In order to realize the maximum possibility of the solution, the first constraint adjustment parameter configured by the loss term weight of the obstacle is preferably 0, so as to delete the constraint of the obstacle from the solving process, so as to perform the path planning decision processing based on the path boundary containing the first constraint adjustment parameter, and obtain the driving path.

[0083] It should be noted that, when the path planning decision processing based on the path boundary containing the first constraint adjustment parameter fails, it means that the loss term weight of the obstacle configured as 0 cannot be normally solved, and therefore, the second constraint adjustment is performed. Specifically, the loss term weight corresponding to all constraint objects in the path boundary is configured as a second constraint adjustment parameter, and the second constraint adjustment parameter is 0, or a small value close to 0, so that when the loss term weight of all constraint objects in the path boundary is configured as 0, it means that in the path planning decision process based on the path boundary containing the second constraint adjustment parameter, the constraint optimization solution is realized or the constraint of all constraint objects is reduced to realize the optimization solution.

[0084] In another embodiment of the present application, in order to further limit and illustrate, after the path planning decision processing based on the path boundary containing the first constraint adjustment parameter, the method further comprises:

[0085] In the case of success of the path planning decision processing based on the path boundary containing the first constraint adjustment parameter, the driving path of the path planning decision processing is determined.

[0086] In the embodiment of the present application, in order to realize effective and fast solution of the driving path, when the path planning decision based on the path boundary containing the first constraint adjustment parameter succeeds, it means that after the loss term weight of the obstacle is configured as 0, a normal solution can be obtained, so as to obtain the driving path based on the path planning decision.

[0087] Similarly, after step 101, if the path boundary based path planning decision can be solved successfully under the condition of path planning decision processing according to the path boundary, it is determined that the normal solution of the path planning decision can be solved based on the path boundary, and thus the driving path of the path planning decision processing is determined, and the embodiment of the present application will not be described here.

[0088] In another embodiment of the present application, in order to further limit and illustrate, the step further comprises:

[0089] Under the condition that the vehicle and the obstacle are expected to collide in the driving path, the path boundary of the vehicle is re-determined, and the path planning decision processing of the vehicle is performed according to the re-determined path boundary.

[0090] In order to accurately perform effective collision avoidance judgment, determine the path boundary of the vehicle, perform path planning decision processing, and obtain the driving path at the latest time, in the embodiment of the present application, when the driving path is obtained, it is detected whether the vehicle and the obstacle collide in the driving path. The obstacle includes dynamic obstacles such as pedestrians and other vehicles, and static obstacles such as walls and road stakes, which are not limited in the embodiment of the present application. When it is determined that the vehicle and the obstacle are expected to collide, the path boundary of the vehicle is re-determined in step 101, and the path planning decision processing is re-performed, so as to realize effective optimization decision.

[0091] In another embodiment of the present application, in order to further limit and illustrate, the path boundary is realized as a constraint strip for planning decision optimization, which greatly reduces the planning error of the path boundary, reduces the collision risk between the vehicle and the obstacle when the turning curvature is too large, and improves the accuracy of the path planning decision based on the path boundary as a constraint condition. Figure 3 As shown in FIG. 8, the step of determining the path boundary of the vehicle comprises:

[0092] 301. In the first coordinate system, the position of the overlapping corner point expected to collide is determined according to the positions of the contour points corresponding to the vehicle and the obstacle in the driving path;

[0093] 302. The contour corner point overlapping line and the decision corner point reference line are determined based on the trajectory normal of the vehicle and the position of the overlapping corner point;

[0094] 303. If the first intersection position between the contour corner point overlapping line and the decision corner point reference line is not on the contour edge, the collision avoidance distance of the vehicle is determined based on the position of the overlapping corner point, the decision corner point reference line, and the position of the contour corner point on the contour corner point overlapping line;

[0095] 304. In the second coordinate system, the path boundary for path planning decision is generated based on the collision avoidance distance.

[0096] The first coordinate system is preferably a Cartesian coordinate system for representing the lateral and longitudinal coordinate positions of the vehicle. The profile point position of the vehicle is the position coordinate of a point corresponding to the vehicle profile in the first coordinate system. The profile point position of the obstacle is the position coordinate of a point corresponding to the obstacle profile in the first coordinate system. Collision detection is performed to determine the overlapping angle point position where the collision is expected to occur. Specifically, since the collision detection is based on the coordinate position, when the collision is determined to occur, the overlapping angle point position is the position coordinate of each angle point in the overlapping part of the profile of the vehicle and the profile of the obstacle in the first coordinate system, i.e., the collision between the vehicle and the obstacle is represented by the profile overlap, as shown in Figure 4 The overlapping angle point position includes at least one main overlapping angle point position Ph and two auxiliary overlapping angle point positions Po1 and Po2. At this time, the overlapping range can be divided by traversing the profile point positions of the vehicle and the obstacle, i.e., the region composed of Ph-Po1-Po2. The overlapping shape is determined based on the range of the overlapping part, and the angle point position of this overlapping shape is determined as the overlapping angle point position, which is not limited in the embodiments of the present application.

[0097] Since the vehicle is automatically driven according to the path trajectory, the trajectory normal of the vehicle driving on the path trajectory can be determined, i.e., a straight line perpendicular to the tangent direction of the path trajectory. The profile angle point overlap line and the decision angle point reference line are determined in combination with the overlapping angle point position. The profile angle point overlap line overlaps the profile edge of the vehicle or the obstacle where the collision is expected to occur. The profile edge is a line segment with two auxiliary overlapping angle points as two end points. That is, for different situations, the profile angle point overlap line can overlap the profile edge Po1-Po2 of the obstacle where the collision is expected to occur, as shown in Figure 5 Or the profile edge Po1-Po2 of the vehicle where the collision is expected to occur, as shown in Figure 6 Wherein, l1 is the trajectory normal, l2 is the profile angle point overlap line, and l3 is the decision angle point reference line. At the same time, the determined decision angle point reference line is parallel to the trajectory normal. Since the decision angle point reference line is constructed based on the overlapping angle point position, and the overlapping angle point position includes the main overlapping angle point position and the auxiliary overlapping angle point position, in order to make the path boundary more accurate, at this time, the decision angle point reference line is fixed based on the main overlapping angle point position, so that the profile angle point overlap line intersects the decision angle point reference line.

[0098] Specifically, in combination with Figure 5 And Figure 6The position relationship of the contour angle point overlapping line l2, the decision angle point reference line l3 in the contour angle point overlapping line l2, and the decision angle point reference line is that the contour angle point overlapping line and the decision angle point reference line are two lines that are not parallel, that is, the contour angle point overlapping line and the decision angle point reference line intersect to obtain a first intersection position. If the first intersection position is not on the contour edge, that is, the included angle between the contour angle point overlapping line and the decision angle point reference line is very small, and the slope of the decision angle point reference line is the same as the normal line of the trajectory, it is indicated that the curvature of the path trajectory is large at this time, and therefore, the collision avoidance distance of the vehicle is determined based on the overlapping angle point position, the decision angle point reference line, and the contour angle point position. The contour angle point position is the position coordinates of two angle points of the contour edge on the contour angle point overlapping line. If the contour angle point overlapping line overlaps and covers the contour edge of the vehicle, the contour angle point position is the angle point position coordinates on the contour of the vehicle. If the contour angle point overlapping line overlaps and covers the contour edge of the obstacle, the contour angle point position is the angle point position coordinates on the contour of the obstacle. Specifically, the angle point in the embodiment of the application is a point of a contour angle, such as two contour angle points Pl1 and Pl2 shown in FIG. 8. Figure 7 The embodiment of the application is not limited in this regard.

[0099] The collision avoidance distance in the embodiment of the application is the distance that the vehicle moves from the current overlapping state to the non-overlapping state in the first coordinate system, so as to realize collision avoidance driving. After the collision avoidance distance is obtained in the first coordinate system, in order to better realize path planning and decision solving, the collision avoidance distance is converted to the second coordinate system to generate a path boundary, that is, the path boundary is used as a constraint condition for optimization solving of path planning and decision. In addition, since the path trajectory is a road to be expected to drive, in order to determine whether the vehicle has lateral avoidance space in path planning and decision, the second coordinate system is a coordinate system constructed by the distance along the trajectory and the distance from the trajectory center line, and is preferably a Frenrt coordinate system. When the collision avoidance distance is determined, it is converted to the Frenrt coordinate system, so as to generate a path boundary based on the collision avoidance distance in the coordinate system. At this time, the path boundary is used for path planning and decision, and the accuracy of path planning is improved.

[0100] It should be noted that the construction and determination of each line in the embodiment of the application can be based on the point position coordinates and the straight line formula For example, the decision angle point reference line is constructed according to the (x, y) coordinates of the main overlapping angle point position and the slope k of the normal line of the trajectory, and the embodiment of the application is not limited in this regard.

[0101] In another embodiment of the application, in order to further limit and illustrate, the step of determining the collision avoidance distance of the vehicle based on the overlapping angle point position, the decision angle point reference line, and the contour angle point position on the contour angle point overlapping line includes:

[0102] If the position of the primary overlapping corner point is the same as the position of the vehicle outline corner point, then a collision avoidance translation line is constructed based on the position of the primary overlapping corner point, the position of the obstacle outline corner point, and the position of the secondary overlapping corner point. The collision avoidance distance of the vehicle is determined according to the position of the primary overlapping corner point, the position of the second intersection between the collision avoidance translation line and the decision corner point reference line. The outline corner point overlapping line overlaps and covers the outline edge of the obstacle where the collision is expected to occur and the position of the obstacle outline corner point.

[0103] If the position of the primary overlapping corner point is the same as the position of the obstacle outline corner point of the obstacle, then a collision avoidance translation line is constructed based on the position of the primary overlapping corner point, the position of the vehicle outline corner point, and the position of the secondary overlapping corner point. The collision avoidance distance of the vehicle is determined according to the position of the primary overlapping corner point, the position of the second intersection between the collision avoidance translation line and the decision corner point reference line. The outline corner point overlapping line overlaps and covers the outline edge of the vehicle where the collision is expected to occur and the position of the vehicle outline corner point.

[0104] To accurately determine vehicle collision avoidance distances, the overlapping edges of the vehicle or obstacle contours covered by the constructed contour corner point overlap lines differ depending on the anticipated collision scenario. Specifically, for example... Figure 5 As shown, when a vehicle is expected to collide with an obstacle, in the area of ​​overlapping collision, the position of the main overlapping corner point is the same as the position of one of the vehicle's contour corner points, indicating that the vehicle enters the obstacle and collides. At this time, the constructed contour corner point overlap line overlaps and covers the contour edge of the obstacle where the collision is expected to occur. Furthermore, the contour corner point overlap line also overlaps and covers the contour corner point position of the obstacle, corresponding to... Figure 5 As shown in l2. Figure 6 As shown, when a vehicle is expected to collide with an obstacle, in the area of ​​overlapping collision, the position of the main overlapping corner point is the same as the position of one of the outline corner points of the obstacle, indicating that the obstacle enters the vehicle and collides with it. At this time, the constructed outline corner point overlap line overlaps and covers the outline edge of the vehicle where the collision is expected. Furthermore, the outline corner point overlap line also overlaps and covers the vehicle's outline corner point position, corresponding to... Figure 6 The l2 shown.

[0105] It should be noted that under different anticipated collision scenarios, the overlap lines of the contour corner points obtained by overlapping the contour edges of the vehicle or obstacle will be different. Collision avoidance translation lines can be constructed based on the positions of the primary overlap corner points, the vehicle contour corner points, the obstacle contour corner points, and the secondary overlap corner points. In this case, the collision avoidance translation line is used to characterize the maximum reference line for the vehicle's collision avoidance movement with the obstacle, such as... Figure 8The l5 shown. Wherein, in order to make the vehicle move in the process of collision avoidance, according to the direction of the trajectory normal of the vehicle, therefore, based on the straight line distance between the second intersection position between the collision avoidance translation line and the decision angle point reference line and the main overlap angle point position as the collision avoidance distance, so that the vehicle avoids collision with the obstacle after trajectory optimization according to this collision avoidance distance as the constraint condition.

[0106] In another embodiment of the present application, in order to further limit and illustrate, the step of constructing a collision avoidance translation line based on the main overlap angle point position, the vehicle contour angle point position or the obstacle contour angle point position, and the auxiliary overlap angle point position, and determining the collision avoidance distance of the vehicle according to the main overlap angle point position and the second intersection position between the collision avoidance translation line and the decision angle point reference line comprises:

[0107] Based on the vector direction of the auxiliary overlap angle point position and the main overlap angle point position relative to the trajectory normal, a target auxiliary overlap angle point position is selected;

[0108] According to the target auxiliary overlap angle point position and the main overlap angle point position, a reference translation line is constructed, and based on the slope of the reference translation line and the vehicle contour angle point position or the obstacle contour angle point position, a collision avoidance translation line is constructed;

[0109] The second intersection position between the collision avoidance translation line and the decision angle point reference line is determined, and the distance between the second intersection position and the main overlap angle point position is determined as the collision avoidance distance.

[0110] In an embodiment of the present application, since the auxiliary overlap angle point position includes the position coordinates of two overlap angle points, in order to construct a reference translation line based on a target auxiliary overlap angle point position and the main overlap angle point position to obtain a collision avoidance translation line, at this time, based on the vector direction of the auxiliary overlap angle point position and the main overlap angle point position relative to the trajectory normal, a target auxiliary overlap angle point position is selected. Wherein, the positions of the two auxiliary overlap angle points are determined in the vector direction with the main overlap angle point position, if at this time, the vector direction of one auxiliary overlap angle point Po1 and the main overlap angle point position Ph relative to the trajectory normal is positive, then the other auxiliary overlap angle point Po2 is selected as the target auxiliary overlap angle point, that is, the position of Po2 is selected as the target auxiliary overlap angle point position, as shown in Figure 8 At this time, the target auxiliary overlap angle point position Po2 and the main overlap angle point position Ph construct a reference translation line. After the reference translation line l4 is constructed, the collision avoidance translation line is constructed based on the slope of the reference translation line l4 and the contour angle point position of the vehicle or the obstacle. In addition, if the vector direction of one overlap angle point Po1 and the main overlap angle point position Ph relative to the trajectory normal is negative, then the way of constructing the collision avoidance translation line is opposite to the positive direction, which is not described in detail in the embodiment of the present application.

[0111] In a specific embodiment scenario, if the main overlapping corner point position Ph (Phx, Phy) is the same as the vehicle contour corner point position, as shown in Figure 8 Po1 (Po1x, Po1y), Po2 (Po2x, Po2y), at this time, the vector direction Proj1 = Proj (<Ph, Po1>, l3) of Po1 and Ph relative to the track normal is positive, then the target auxiliary overlapping corner point position Po2 is selected to construct the reference translation line l4 with the main overlapping corner point position Ph, at this time, the direction of the decision corner reference line l3 is the same as the direction of the track normal l1. Then, based on the slope of the reference translation line l4 and one of the obstacle contour corner point positions Pl1 (Pl1x, Pl1y), Pl2 (Pl2x, Pl2y), an obstacle avoidance translation line is constructed, preferably an obstacle avoidance translation line l5 is constructed based on the slope of the reference translation line l4 and the obstacle contour corner point position Pl1, at this time,

[0112] k_l4 = (Po2y - Phy) / (Po2x - Phx), that is, l5: y = k_l4 * (x - Pl1x) + Pl1y.

[0113] In a specific embodiment scenario, if the main overlapping corner point position Ph (Phx, Phy) is the same as the vehicle contour corner point position, as shown in Figure 9 Po1 (Po1x, Po1y), Po2 (Po2x, Po2y), at this time, the vector direction Proj1 = Proj (<Ph, Po1>, l3) of Po1 and Ph relative to the track normal is positive, then the target auxiliary overlapping corner point position Po1 is selected to construct the reference translation line l4 with the main overlapping corner point position Ph, at this time, the direction of the decision corner reference line l3 is the same as the direction of the track normal l1. Then, based on the slope of the reference translation line l4 and one of the vehicle contour corner point positions Pl1 (Pl1x, Pl1y), Pl2 (Pl2x, Pl2y), an obstacle avoidance translation line is constructed, preferably an obstacle avoidance translation line l5 is constructed based on the slope of the reference translation line l4 and the vehicle contour corner point position Pl2, that is, l5: y = k_l4 * (x - Pl2x) + Pl2y.

[0114] It should be noted that, as shown in Figure 8 and as shown in Figure 9 The obstacle avoidance translation line l5 and the decision corner reference line l3 both intersect at a second intersection position, that is, Pi (Pix, Piy), at this time, in order to make the vehicle and the obstacle completely avoid collision during path planning, the distance between the second intersection position Pi and the main overlapping corner point position Ph is determined as the obstacle avoidance distance, that is, adjust_distance = ||<Pi, Ph>||.

[0115] In another embodiment of the present application, in order to further define and illustrate, the step of constructing the collision avoidance translation line based on the slope of the reference translation line and the vehicle profile corner point position or the obstacle profile corner point position comprises:

[0116] Based on the vector direction of the vehicle profile corner point position or the obstacle profile corner point position and the main overlap corner point position relative to the trajectory normal line, a target vehicle profile corner point position or a target obstacle profile corner point position is selected, and a collision avoidance translation line is constructed based on the slope of the reference translation line and the target vehicle profile corner point position or the target obstacle profile corner point position.

[0117] Since the profile corner point overlap line covers the vehicle profile corner point position or the obstacle profile corner point position, the vehicle profile corner point position or the obstacle profile corner point position respectively includes the position coordinates of two profile corner points, such as Pl1, Pl2, therefore, in order to accurately construct the collision avoidance translation line, when constructing the collision avoidance translation line, specifically, based on the vector direction of the vehicle profile corner point position or the obstacle profile corner point position and the main overlap corner point position relative to the trajectory normal line, a target vehicle profile corner point position or a target obstacle profile corner point position is selected, and a collision avoidance translation line is constructed based on the slope of the reference translation line and the target vehicle profile corner point position or the target obstacle profile corner point position.

[0118] In the embodiment scenario as shown in Figure 8 the main overlap corner point position Ph (Phx, Phy) is the same as the vehicle profile corner point position of the vehicle, the obstacle profile corner point positions Pl1, Pl2, at this time, if the vector direction Proj1' = Proj (<Ph, Pl1>, l3) of Pl1 and Ph relative to the trajectory normal line is positive, the target obstacle profile corner point position Pl1 is selected, and a collision avoidance translation line l5 is constructed based on the slope of the reference translation line l4 and the target obstacle profile corner point position Pl1, at this time, l5: y = k_l4 * (x-Pl1x) + P11y.

[0119] Similarly, in the embodiment scenario as shown in Figure 9 the main overlap corner point position Ph (Phx, Phy) is the same as the obstacle profile corner point position of the obstacle, the vehicle profile corner point positions Pl1, Pl2, at this time, if the vector direction Proj1' = Proj (<Ph, Pl1>, l3) of Pl1 and Ph relative to the trajectory normal line is positive, the target vehicle profile corner point position Pl2 is selected, and a collision avoidance translation line l5 is constructed based on the slope of the reference translation line l4 and the target vehicle profile corner point position Pl2, l5: y = k_l4 * (x-Pl2x) + P12y.

[0120] In another embodiment of the present application, in order to further define and illustrate, the step comprises determining a contour corner overlapping line based on the trajectory normal of the vehicle and the overlapping corner position, and a decision corner reference line comprises:

[0121] The contour corner overlapping line is constructed according to two auxiliary overlapping corner positions of the overlapping corner position, and the decision corner reference line is constructed according to the slope of the trajectory normal based on a main overlapping corner position of the overlapping corner position.

[0122] Since the vehicle is automatically driven according to the path trajectory, and the overlapping corner position comprises a main overlapping corner position and auxiliary overlapping corner positions, in order to accurately construct the contour corner overlapping line and the decision corner reference line as shown in Figure 5 、 6 , specifically, the contour corner overlapping line is constructed according to two auxiliary overlapping corner positions of the overlapping corner position, such as l2: y2=k2x+b2, which can be calculated in combination with the two auxiliary overlapping corner positions Po1 (Po1x, Pol1y), Po1 (Po2x, Pol2y). At the same time, the decision corner reference line l3 is constructed based on the slope k1 of the trajectory normal and the main overlapping corner position Ph (Phx, Phy) of the overlapping corner position, l3: y3=k1 (x-Phx)+Phy. In addition, the trajectory normal l1 can be represented by l1: y1=k1x+b1, which is not specifically limited in the embodiment of the present application.

[0123] In another embodiment of the present application, after the contour corner overlapping line and the decision corner reference line are determined based on the trajectory normal of the vehicle and the overlapping corner position, the method further comprises:

[0124] If the first intersection position between the contour corner overlapping line and the decision corner reference line is on the contour edge, the distance between the first intersection position and the main overlapping corner position of the overlapping corner position is determined as the collision avoidance distance.

[0125] In the embodiment of the present application, according to the positional relationship of the contour corner overlapping line l2 and the decision corner reference line l3, it is determined that the contour corner overlapping line and the decision corner reference line are two non-parallel lines, that is, the contour corner overlapping line and the decision corner reference line will intersect to obtain the first intersection position. If the first intersection position is on the contour edge, that is, the included angle between the contour corner overlapping line and the decision corner reference line is large, and the slope of the decision corner reference line is the same as the trajectory normal, it indicates that the curvature of the path trajectory is small at this time, and therefore the distance between the first intersection position and the main overlapping corner position of the overlapping corner position can be determined as the collision avoidance distance. As shown in Figure 7As shown, if the first intersection position P1' (P1'x, P1'y) of l2 and l3 is located on the contour edge of the obstacle, then directly calculate the distance between the first intersection position P1' and the main overlap angle point position Ph (Phx, Phy) as the collision avoidance distance, that is

[0126] adjust_distance = || <P1', Ph> ||.

[0127] In another embodiment of the present application, in the second coordinate system, generating a path boundary for path planning decision based on the collision avoidance distance comprises:

[0128] Converting the collision avoidance distance to a collision avoidance length in the second coordinate system;

[0129] Based on the lateral distance between the target trajectory point in the path trajectory in the second coordinate system and the vehicle, coupling the collision avoidance length and the decision length to obtain the path boundary for path planning decision.

[0130] In order to make path planning decision as a constraint condition in the form of path boundary, and improve the accuracy of path planning, the path boundary is generated based on the collision avoidance distance in the second coordinate system. Specifically, first, the collision avoidance distance obtained in the first coordinate system is converted to the second coordinate system, such as converting the adjust_daistance obtained in the Cartesian coordinate system to the adjusted_daistance in the Frenet coordinate system, so as to couple the collision avoidance length and the decision length by the lateral distance between the target trajectory point in the second coordinate system and the vehicle, and obtain the path boundary. Wherein, the target trajectory point is a trajectory point on the path trajectory, and one trajectory frame can contain 20 trajectory points within 0-4s, and the current trajectory point is i, so as to determine the position of i+1 trajectory points in the Frenet coordinate system when making path planning decision, so as to determine the lateral distance between each target trajectory point and the vehicle. At this time, the center point of the vehicle can be taken as the reference position to determine the lateral distance, so as to constrain the vehicle relative to the path trajectory in the lateral direction, which is not limited in the embodiment of the application. In addition, after determining the lateral distance l between the target trajectory point and the vehicle, the lateral distance and the length of the obstacle are compared to couple the collision avoidance length and the decision length according to the comparison result to obtain the path boundary as a constraint condition. Specifically, if the lateral distance of the i+1 point is less than the length of the obstacle, then path.upper_bound=traj(i).l-adjusted_distance; if the lateral distance of the i+1 point is greater than the length of the obstacle, then path.lower_bound=traj(i).l+adjusted_distance; wherein, traj(i).l represents the decision length corresponding to the i point, so as to couple the collision avoidance length adjusted_distance in different cases to obtain the path boundary path bound, and path.upper_bound and path.lower_bound are the upper limit and the lower limit of the path planning of the next trajectory point as the constraint condition of the path boundary.

[0131] In another embodiment of the application, in the first coordinate system, before determining the position of the overlapping angle point where the collision is expected to occur according to the positions of the contour points corresponding to the vehicle and the obstacle in the path trajectory, the method further comprises:

[0132] In the second coordinate system, the path trajectory of the vehicle in the predicted trajectory frame is obtained, and the path trajectory is converted to the first coordinate system;

[0133] In the first coordinate system, based on the vehicle contour of the vehicle, the position of the vehicle contour point of the vehicle at the target trajectory point in the path trajectory is determined;

[0134] In the first coordinate system, based on the perception system and / or the path map, an obstacle contour point position of at least one of the obstacles matching the target trajectory point in the path trajectory is acquired.

[0135] In order to determine the path boundary in the first coordinate system and improve the optimization accuracy of the path boundary, the path trajectory of the vehicle in the prediction trajectory frame is first acquired in the second coordinate system, so as to convert the path trajectory to the first coordinate system. The prediction trajectory frame can include all path trajectory points corresponding to the current moment and the future 4 seconds, and preferably 20 path trajectory points, so as to form a path trajectory. In the embodiment of the present application, the std: <trajectorypoints>The trajectory application carries out vehicle model acquisition, determines the path trajectory of the vehicle, and converts to the first coordinate system. Meanwhile, the obstacle decision maker determines the vehicle contour point position corresponding to each target trajectory point in the path trajectory according to the vehicle contour of the vehicle in the first coordinate, so as to determine the vehicle contour point position of the vehicle in the first coordinate system based on the vehicle contour. At this time, the vehicle contour point position includes a plurality of vehicle contour edge point positions and four vehicle contour corner point positions, which are not limited in the embodiment of the application.

[0136] It should be noted that, since the prediction trajectory frame in the embodiment of the application can include 20 target trajectory points, collision detection is carried out on the vehicle at each point, therefore, in the first coordinate system, the obstacle contour point position of at least one of the obstacles matched with the target trajectory point in the path trajectory is acquired based on the perception system and / or the path map, so as to detect whether the vehicle collides with the obstacle based on the obstacle contour point position. The obstacle contour point position includes a plurality of obstacle contour edge point positions and four obstacle contour corner point positions, which are not limited in the embodiment of the application. Meanwhile, collision detection between the obstacle and the vehicle in the first coordinate system can also be carried out by the HasOverlap function with the coverage area calculation function, that is, whether there is an overlapping coverage area is calculated based on the obstacle contour point position and the vehicle contour point position in the first coordinate system, so as to complete the collision detection, if collision is expected to occur at each target trajectory point, the overlapping corner point position corresponding to each target trajectory point is calculated, so as to execute the method of steps 301-304 in the embodiment of the application.

[0137] In another embodiment of the application, the vehicle contour point position of the vehicle at the target trajectory point in the path trajectory is determined based on the vehicle contour of the vehicle, which can be specifically: the heading angle of the target trajectory point in the path trajectory is calculated, and the vehicle contour point position of the vehicle is determined according to the heading angle and the vehicle contour.

[0138] Since the vehicle is driven along the path trajectory by controlling the center point of the vehicle during driving, i.e., the position of the vehicle during driving is changed in real time, in order to accurately determine the position of the vehicle contour point, when the position of the vehicle contour point is determined based on the vehicle contour in the first coordinate system, specifically, the heading angle of each target trajectory point is calculated first, and based on the heading angle and the vehicle contour (including the length and width of the vehicle), the vehicle contour point position of the vehicle is drawn in the first coordinate system, the vehicle contour point position includes a plurality of vehicle contour edge point positions and four vehicle contour corner point positions. Wherein, the heading angle is based on the tangent direction of the path trajectory, i.e., the heading angle heading angle: atan2 (x_{i}-x_{i-1}, y_{i}-y_{i-1}), i=1,...n, n=20.

[0139] In another embodiment of the present application, before determining the contour corner point overlap line based on the trajectory normal line of the vehicle, the overlap angle point position, and the decision corner point reference line, the method further comprises:

[0140] Obtaining the center point position of the vehicle and determining the vertical direction of the path trajectory;

[0141] Constructing the trajectory normal line of the vehicle based on the vertical direction and the center point position.

[0142] Since the center point of the vehicle is automatically driven along the path trajectory in the embodiment of the present application, in order to make the vehicle plan according to the lateral distance between the driving path during collision avoidance, the trajectory normal line is constructed in advance. Specifically, first, the center point position corresponding to each target trajectory point at the moment is obtained, at this time, the geometric calculation can be performed based on the vehicle contour in the first coordinate system, which is not limited in the embodiment of the present application. At the same time, since the heading angle is the tangent direction of the path trajectory, at this time, the perpendicular line can be made based on the tangent direction of the path trajectory, and the vertical direction is obtained, i.e., the direction of the trajectory normal line, so as to construct the trajectory normal line of the vehicle based on the vertical direction through the center point position, such as l1 in Figure 5 Figure 6 At this time, the direction of the trajectory normal line is towards the path trajectory, so as to determine the vector direction corresponding to the projection Proj1 and the projection Proj1'.

[0143] ​The embodiment of the present application provides a driving path planning method, compared with the prior art, the embodiment of the present application determines the path boundary of a vehicle, and carries out path planning decision processing of the vehicle according to the path boundary, the path boundary is generated according to the collision avoidance distance of the trajectory normal of the vehicle and the overlapping angle point position of the vehicle and an obstacle expected to collide; in the condition that the path planning decision processing according to the path boundary fails, the path boundary is adjusted and constrained, and the path planning decision processing is carried out based on the path boundary after the constraint adjustment; in the condition that the path planning decision processing based on the path boundary after the constraint adjustment fails, the reference driving trajectory of the vehicle is acquired, and the driving path of the path planning decision is determined according to the reference driving trajectory, the optimization solution result is obtained in the manner of reducing the constraint limitation, the solving effectiveness of the path planning decision is improved, and therefore the effectiveness of the driving path planning is improved.

[0144] Further, as the implementation of the method shown in the above Figure 1 The embodiment of the present application provides a driving path planning device, as shown in the above Figure 10 The device comprises:

[0145] A determination module 41 is configured to determine the path boundary of a vehicle, and carry out path planning decision processing of the vehicle according to the path boundary, the path boundary is generated according to the collision avoidance distance of the trajectory normal of the vehicle and the overlapping angle point position of the vehicle and an obstacle expected to collide;

[0146] An adjustment module 42 is configured to, in the condition that the path planning decision processing according to the path boundary fails, adjust and constrain the path boundary, and carry out path planning decision processing based on the path boundary after the constraint adjustment;

[0147] An acquisition module 43 is configured to, in the condition that the path planning decision processing based on the path boundary after the constraint adjustment fails, acquire the reference driving trajectory of the vehicle, and determine the driving path of the path planning decision according to the reference driving trajectory.

[0148] Further, the adjustment module comprises:

[0149] A first configuration unit is configured to configure the loss term weight corresponding to the obstacle in the path boundary as a first constraint adjustment parameter, and carry out path planning decision processing based on the path boundary containing the first constraint adjustment parameter;

[0150] The second configuration unit is configured to, in a case where the path planning decision processing based on the path boundary containing the first constraint adjustment parameter fails, configure loss term weights corresponding to all constraint objects in the path boundary as a second constraint adjustment parameter, and perform path planning decision processing based on the path boundary containing the second constraint adjustment parameter.

[0151] Further, the determination module is further configured to, in a case where the path planning decision processing based on the path boundary containing the first constraint adjustment parameter succeeds, determine a driving path of the path planning decision processing.

[0152] Further, the determination module is further configured to, in a case where the vehicle is detected to be expected to collide with the obstacle in the driving path, re-determine a path boundary of the vehicle, and perform path planning decision processing of the vehicle based on the re-determined path boundary.

[0153] Further, the determination module comprises:

[0154] A first determination unit is configured to, in a first coordinate system, determine a position of an overlap corner point expected to collide based on positions of contour points corresponding to the vehicle and the obstacle in the driving path, respectively.

[0155] A second determination unit is configured to determine a contour corner point overlap line based on a trajectory normal of the vehicle, the position of the overlap corner point, and a decision corner point reference line, the decision corner point reference line being parallel to the trajectory normal, and the contour corner point overlap line overlapping and covering a contour edge of the vehicle or the obstacle expected to collide.

[0156] A third determination unit is configured to, if a first intersection position between the contour corner point overlap line and the decision corner point reference line is not on the contour edge, determine a collision avoidance distance of the vehicle based on the position of the overlap corner point, the decision corner point reference line, and a contour corner point position on the contour corner point overlap line.

[0157] A generation unit is configured to, in a second coordinate system, generate a path boundary for path planning decision based on the collision avoidance distance.

[0158] Further, the overlapping corner point position comprises at least one main overlapping corner point position and two auxiliary overlapping corner point positions, the third determining unit is specifically configured to: if the main overlapping corner point position is the same as the vehicle contour corner point position of the vehicle, construct an avoidance translation line based on the main overlapping corner point position, the obstacle contour corner point position and the auxiliary overlapping corner point position, and determine the avoidance distance of the vehicle according to a second intersection position between the main overlapping corner point position, the avoidance translation line and the decision corner point reference line, the contour corner point overlapping line being overlapped on the contour edge of the obstacle expected to collide and the obstacle contour corner point position; if the main overlapping corner point position is the same as the obstacle contour corner point position of the obstacle, construct an avoidance translation line based on the main overlapping corner point position, the vehicle contour corner point position and the auxiliary overlapping corner point position, and determine the avoidance distance of the vehicle according to a second intersection position between the main overlapping corner point position, the avoidance translation line and the decision corner point reference line, the contour corner point overlapping line being overlapped on the contour edge of the vehicle expected to collide and the vehicle contour corner point position.

[0159] Further, the third determining unit is specifically further configured to: select a target auxiliary overlapping corner point position based on the vector direction of the auxiliary overlapping corner point position and the main overlapping corner point position relative to the trajectory normal line; construct a reference translation line according to the target auxiliary overlapping corner point position and the main overlapping corner point position, and construct an avoidance translation line based on the slope of the reference translation line and the vehicle contour corner point position or the obstacle contour corner point position; determine a second intersection position between the avoidance translation line and the decision corner point reference line, and determine the avoidance distance as the distance between the second intersection position and the main overlapping corner point position.

[0160] Further, the third determining unit is specifically further configured to: select a target vehicle contour corner point position or a target obstacle contour corner point position based on the vector direction of the vehicle contour corner point position or the obstacle contour corner point position relative to the main overlapping corner point position, and construct an avoidance translation line according to the slope of the reference translation line and the target vehicle contour corner point position or the target obstacle contour corner point position.

[0161] Further, the second determining unit is specifically configured to: construct a contour corner point overlapping line according to the two auxiliary overlapping corner point positions in the overlapping corner point position, and construct a decision corner point reference line according to the slope of the trajectory normal line based on the main overlapping corner point position in the overlapping corner point position.

[0162] This invention provides a driving path planning device. Compared with the prior art, this invention determines the path boundary of a vehicle and performs path planning decision processing based on the path boundary. The path boundary is generated based on the trajectory normal of the vehicle and the collision avoidance distance determined by the overlapping corner position of the expected collision between the vehicle and the obstacle. If the path planning decision processing based on the path boundary fails, the path boundary is constrained and adjusted, and path planning decision processing is performed based on the constrained path boundary. If the path planning decision processing based on the constrained path boundary fails again, a reference driving trajectory of the vehicle is obtained, and the driving path for the path planning decision is determined based on the reference driving trajectory. This achieves optimized solution results by reducing constraints, improves the effectiveness of path planning decision solving, and thus improves the effectiveness of driving path planning.

[0163] Furthermore, as an implementation of the above method, embodiments of the present invention provide a vehicle including the aforementioned driving path planning device.

[0164] According to one embodiment of the present invention, a storage medium is provided, the storage medium storing at least one executable instruction, the computer-executable instruction being able to execute the driving path planning method in any of the above method embodiments.

[0165] Figure 11 The diagram shows a structural schematic of a terminal according to an embodiment of the present invention. The specific implementation of the terminal is not limited by the specific embodiments of the present invention.

[0166] like Figure 11 As shown, the terminal may include: a processor 502, a communications interface 504, a memory 506, and a communications bus 508.

[0167] The processor 502, communication interface 504, and memory 506 communicate with each other via communication bus 508.

[0168] Communication interface 504 is used to communicate with other network elements such as clients or other servers.

[0169] The processor 502 is used to execute program 510, which can specifically execute the relevant steps in the above-described driving path planning method embodiment.

[0170] Specifically, program 510 may include program code that includes computer operation instructions.

[0171] The processor 502 can be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to perform the embodiments of the present application. The terminal can include one or more processors of the same type, such as one or more CPUs, or processors of different types, such as one or more CPUs and one or more ASICs.

[0172] The memory 506 is configured to store a program 510. The memory 506 can include a high-speed RAM memory, and can further include a non-volatile memory, such as at least one disk memory.

[0173] The program 510 can be specifically configured to cause the processor 502 to perform the following operations:

[0174] determining a path boundary of the vehicle, and making a path planning decision of the vehicle according to the path boundary, the path boundary being generated according to a collision avoidance distance determined according to a trajectory normal of the vehicle and an overlapping angle point position at which the vehicle is expected to collide with an obstacle;

[0175] under a condition that the path planning decision fails according to the path boundary, adjusting the path boundary by constraint, and making a path planning decision based on the path boundary adjusted by constraint;

[0176] under a condition that the path planning decision fails based on the path boundary adjusted by constraint, obtaining a reference driving trajectory of the vehicle, and determining a driving path of the path planning decision according to the reference driving trajectory.

[0177] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present application can be realized by general computing devices, which can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be realized by program codes executable by computing devices, so that they can be stored in storage devices and executed by computing devices, and in some cases, the steps shown or described can be executed in different order, or they can be manufactured into individual integrated circuit modules, or multiple modules or steps can be manufactured into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.

[0178] The above merely provides the preferred embodiments of the present application, and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the principles and technical scope of the present application shall fall into the scope of the present application.< / trajectorypoints>

Claims

1. A method for planning a driving route, characterized in that, include: The path boundary of the vehicle is determined, and the path planning decision is made based on the path boundary. The path boundary is generated based on the trajectory normal of the vehicle and the collision avoidance distance determined by the overlapping corner point position of the vehicle and the expected collision between the vehicle and the obstacle. If path planning decision processing fails based on the path boundary, the path boundary is constrained and adjusted, and path planning decision processing is performed based on the constrained and adjusted path boundary. If the path planning decision processing based on the constrained adjusted path boundary fails, the reference driving trajectory of the vehicle is obtained, and the driving path for the path planning decision is determined based on the reference driving trajectory. The determination of the vehicle's path boundaries includes: In the first coordinate system, the position of the overlapping corner point where the expected collision is to occur is determined based on the position of the contour points corresponding to the vehicle and the obstacle in the driving path; Based on the trajectory normal of the vehicle, the overlapping corner point position, the contour corner point overlap line and the decision corner point reference line are determined. The decision corner point reference line is parallel to the trajectory normal, and the contour corner point overlap line overlaps and covers the contour edge where the vehicle or the obstacle is expected to collide. If the first intersection point between the overlapping contour corner point line and the decision corner point reference line is not located on the contour edge, the collision avoidance distance of the vehicle is determined based on the overlapping corner point position, the decision corner point reference line, and the contour corner point position located on the overlapping contour corner point line. In the second coordinate system, a path boundary is generated based on the collision avoidance distance for path planning decisions.

2. The method according to claim 1, characterized in that, The process of adjusting the constraints on the path boundaries and making path planning decisions based on the adjusted path boundaries includes: The weights of the loss terms corresponding to the obstacles in the path boundary are configured as first constraint adjustment parameters, and path planning decision processing is performed based on the path boundary containing the first constraint adjustment parameters. If the path planning decision processing based on the path boundary containing the first constraint adjustment parameter fails, the weights of the loss terms corresponding to all constraint objects in the path boundary are configured as the second constraint adjustment parameter, and the path planning decision processing is performed based on the path boundary containing the second constraint adjustment parameter.

3. The method according to claim 2, characterized in that, After performing path planning decision processing based on the path boundary including the first constraint adjustment parameter, the method further includes: Under the condition that the path planning decision processing is successful based on the path boundary including the first constraint adjustment parameter, the driving path of the path planning decision processing is determined.

4. The method according to claim 1, characterized in that, The method further includes: If a collision between the vehicle and the obstacle is expected to occur during the driving path, the path boundary of the vehicle is redefined, and the path planning decision of the vehicle is performed based on the redefined path boundary.

5. The method according to claim 1, characterized in that, The overlapping corner point positions include at least one primary overlapping corner point position and two secondary overlapping corner point positions. Determining the collision avoidance distance of the vehicle based on the overlapping corner point positions, the decision corner point reference line, and the contour corner point positions located on the overlapping contour corner point line includes: If the position of the primary overlapping corner point is the same as the position of the vehicle outline corner point, then a collision avoidance translation line is constructed based on the position of the primary overlapping corner point, the position of the obstacle outline corner point, and the position of the secondary overlapping corner point. The collision avoidance distance of the vehicle is determined according to the position of the primary overlapping corner point, the position of the second intersection between the collision avoidance translation line and the decision corner point reference line. The outline corner point overlapping line overlaps and covers the outline edge of the obstacle where the collision is expected to occur and the position of the obstacle outline corner point. If the position of the primary overlapping corner point is the same as the position of the obstacle outline corner point of the obstacle, then a collision avoidance translation line is constructed based on the position of the primary overlapping corner point, the position of the vehicle outline corner point, and the position of the secondary overlapping corner point. The collision avoidance distance of the vehicle is determined according to the position of the primary overlapping corner point, the position of the second intersection between the collision avoidance translation line and the decision corner point reference line. The outline corner point overlapping line overlaps and covers the outline edge of the vehicle where the collision is expected to occur and the position of the vehicle outline corner point.

6. The method according to claim 5, characterized in that, The process of constructing a collision avoidance translation line based on the main overlapping corner point position, the vehicle outline corner point position or the obstacle outline corner point position, and the secondary overlapping corner point position, and determining the vehicle's collision avoidance distance based on the main overlapping corner point position, the second intersection point between the collision avoidance translation line and the decision corner point reference line, includes: Based on the vector directions of the secondary overlapping corner point position and the primary overlapping corner point position relative to the trajectory normal, the target secondary overlapping corner point position is selected. A reference translation line is constructed based on the target secondary overlapping corner point position and the main overlapping corner point position, and a collision avoidance translation line is constructed based on the slope of the reference translation line and the vehicle outline corner point position or the obstacle outline corner point position. Determine the position of the second intersection point between the collision avoidance translation line and the decision corner reference line, and determine the distance between the second intersection point position and the main overlapping corner point position as the collision avoidance distance.

7. The method according to claim 6, characterized in that, The construction of the collision avoidance translation line based on the slope of the reference translation line and the position of the vehicle contour corner or the position of the obstacle contour corner includes: Based on the vector direction of the vehicle outline corner position or the obstacle outline corner position and the main overlapping corner position relative to the trajectory normal, the target vehicle outline corner position or the target obstacle outline corner position is selected, and a collision avoidance translation line is constructed according to the slope of the reference translation line and the target vehicle outline corner position or the target obstacle outline corner position.

8. The method according to claim 1, characterized in that, The determination of the contour corner overlap line based on the vehicle's trajectory normal, the overlapping corner point position, and the decision corner point reference line includes: Contour corner overlap lines are constructed based on the positions of two secondary overlapping corner points among the overlapping corner point positions, and decision corner reference lines are constructed based on the slope of the trajectory normal line according to the position of the primary overlapping corner point among the overlapping corner point positions.

9. A driving path planning device, characterized in that, include: The determination module is used to determine the path boundary of the vehicle and perform path planning decision processing for the vehicle based on the path boundary. The path boundary is generated based on the trajectory normal of the vehicle and the collision avoidance distance determined by the overlapping corner position of the vehicle and the expected collision between the vehicle and the obstacle. The adjustment module is used to adjust the constraints of the path boundary when the path planning decision processing based on the path boundary fails, and to perform path planning decision processing based on the constrained path boundary. The acquisition module is used to acquire the reference driving trajectory of the vehicle when the path planning decision processing based on the path boundary after constraint adjustment fails, and to determine the driving path for path planning decision based on the reference driving trajectory. The determining module includes: The first determining unit is used to determine the position of the overlapping corner point where the collision is expected to occur, based on the positions of the contour points corresponding to the vehicle and the obstacle in the driving path, in the first coordinate system. The second determining unit is used to determine the contour corner overlap line and the decision corner reference line based on the trajectory normal of the vehicle and the position of the overlap corner point. The decision corner reference line is parallel to the trajectory normal, and the contour corner overlap line overlaps and covers the contour edge where the vehicle or the obstacle is expected to collide. The third determining unit is used to determine the collision avoidance distance of the vehicle based on the overlapping corner position, the decision corner reference line, and the contour corner position on the overlapping corner line if the first intersection point between the overlapping contour corner line and the decision corner reference line is not on the contour edge. The generation unit is used to generate a path boundary for path planning decisions based on the collision avoidance distance in the second coordinate system.

10. A vehicle, characterized in that, Includes the driving path planning device as described in claim 9.

11. A storage medium storing at least one executable instruction that causes a processor to perform the driving path planning method as described in any one of claims 1-8.

12. A terminal, comprising: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to perform the driving path planning method as described in any one of claims 1-8.

Images