Trajectory planning method, device, equipment and storage medium
The nonlinear proposition of trajectory planning is simplified by a method based on tunneling modeling, and the problem of high computational complexity in the prior art is solved, and the efficiency and accuracy of trajectory planning are improved.
Patent Information
- Application Number
- CN202010414249.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-05-15
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2040-05-15
AI Technical Summary
In the unmanned autonomous driving scenario, the existing trajectory planning method takes into account the information of all obstacles, resulting in high computational complexity and reduces the efficiency of trajectory planning.
The collision avoidance constraints are determined based on tunneling modeling, a simple nonlinear proposition for target trajectory planning is constructed, and the target planning trajectory of the target vehicle is solved by minimizing the trajectory cost function.
The complexity of collision avoidance constraints is reduced, redundant calculations are avoided, and trajectory planning efficiency and accuracy are improved.
Smart Images

Figure CN113671941B_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present invention relate to computer technology, and in particular to a trajectory planning method, apparatus, device and storage medium. Background Art
[0002] With the rapid development of computer technology, vehicles can achieve unmanned autonomous driving on structured roads. In unmanned autonomous driving scenarios, it is usually necessary to make trajectory decisions (i.e., operation decisions) for the vehicle to determine which side the vehicle should go around obstacles from, and whether to take the lead or give way. After making trajectory decisions for the vehicle, trajectory planning is required to smooth the rough driving trajectory decided, and obtain a more accurate planned trajectory.
[0003] At present, the existing trajectory planning method is to plan according to the information of all obstacles in the driving scene to ensure that the vehicle will not collide with obstacles when driving along the planned trajectory.
[0004] However, in the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art:
[0005] Due to the limited length of the vehicle, it is unlikely that the vehicle will collide with all obstacles at all times. Therefore, the existing trajectory planning method considers the information of all obstacles at all times, which will lead to a large amount of redundant calculations and greatly reduce the efficiency of trajectory planning. Summary of the invention
[0006] Embodiments of the present invention provide a trajectory planning method, apparatus, device and storage medium to improve trajectory planning efficiency.
[0007] In a first aspect, an embodiment of the present invention provides a trajectory planning method, including:
[0008] Obtaining initial trajectory information obtained by making trajectory decisions for the target vehicle;
[0009] Constructing a nonlinear proposition of target trajectory planning corresponding to the target vehicle, wherein the nonlinear proposition of target trajectory planning includes: a trajectory cost function and driving constraints, and the driving constraints include: a collision avoidance constraint determined based on a tunneling modeling method;
[0010] The target trajectory planning nonlinear proposition is solved according to the initial trajectory information to determine the target planning trajectory corresponding to the target vehicle.
[0011] In a second aspect, an embodiment of the present invention further provides a trajectory planning device, comprising:
[0012] An initial trajectory information acquisition module is used to obtain initial trajectory information obtained by making trajectory decisions on the target vehicle;
[0013] A trajectory planning nonlinear proposition construction module is used to construct a target trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the target trajectory planning nonlinear proposition includes: a trajectory cost function and a driving constraint condition, and the driving constraint condition includes: a collision avoidance constraint condition determined based on a tunneling modeling method;
[0014] The target planning trajectory determination module is used to solve the target trajectory planning nonlinear proposition according to the initial trajectory information to determine the target planning trajectory corresponding to the target vehicle.
[0015] In a third aspect, an embodiment of the present invention further provides a device, the device comprising:
[0016] one or more processors;
[0017] A memory for storing one or more programs;
[0018] When the one or more programs are executed by the one or more processors, the one or more processors implement the trajectory planning method as described in any embodiment of the present invention.
[0019] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the trajectory planning method as described in any embodiment of the present invention.
[0020] The embodiments of the above invention have the following advantages or beneficial effects:
[0021] The collision avoidance constraints are determined based on the tunnel modeling method, that is, the drivable tunnel area of the target vehicle at each moment is obtained, so that there is no need to consider obstacles that will not affect the driving of the target vehicle during the trajectory planning process, which reduces the complexity of the collision avoidance constraints, thereby constructing a simpler target trajectory planning nonlinear proposition, and the target trajectory planning nonlinear proposition can be quickly solved by minimizing the trajectory cost function to obtain the target planning trajectory corresponding to the target vehicle, thereby avoiding redundant calculations and improving the trajectory planning efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1 is a flow chart of a trajectory planning method provided by Embodiment 1 of the present invention;
[0023] Figure 2 This is an example of a driving scenario involved in the first embodiment of the present invention;
[0024] Figure 3It is an example of a two-degree-of-freedom vehicle kinematic model involved in the first embodiment of the present invention;
[0025] Figure 4 This is an example of a local tunnel area involved in the first embodiment of the present invention;
[0026] Figure 5 is a flow chart of a trajectory planning method provided by Embodiment 2 of the present invention;
[0027] Figure 6 This is an example of a construction process of a local tunnel area involved in the second embodiment of the present invention;
[0028] Figure 7 is a flow chart of a trajectory planning method provided by Embodiment 3 of the present invention;
[0029] Figure 8 is a structural schematic diagram of a trajectory planning device provided by Embodiment 4 of the present invention;
[0030] Fig. 9 It is a structural schematic diagram of a device provided in Embodiment 5 of the present invention. DETAILED DESCRIPTION
[0031] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.
[0032] Embodiment 1
[0033] Figure 1 This is a flow chart of a trajectory planning method provided in Embodiment 1 of the present invention. This embodiment is applicable to the case where an autonomous driving vehicle performs trajectory planning, and is particularly applicable to the case where an autonomous driving vehicle performs trajectory planning on a structured road. The method can be executed by a trajectory planning device, which can be implemented by software and / or hardware and integrated into a device with an autonomous driving function, such as any type of vehicle. Figure 1 As shown, the method specifically comprises the following steps:
[0034] S110: Acquire initial trajectory information obtained by performing trajectory decision on the target vehicle.
[0035] Among them, the target vehicle may refer to any type of autonomous driving vehicle that needs trajectory planning on the road. The road may refer to a structured road. A structured road may refer to a road with driving rules on which the target vehicle is to travel. Exemplarily, a structured road may be a road with guide lines to indicate the direction of travel, such as an urban road. Trajectory decision may refer to a decision trajectory for deciding the target vehicle on the road, that is, determining which side the target vehicle will bypass the obstacle from when encountering an obstacle and deciding whether to take the lead or give way. The initial trajectory information may refer to the trajectory information corresponding to the decision trajectory obtained after the trajectory decision. For example, the initial trajectory information may include, but is not limited to: the driving path information and driving speed information of the decided target vehicle.
[0036] Specifically, based on the information of each static obstacle and each dynamic obstacle in the structured road and the target vehicle's own information, a trajectory decision can be made for the target vehicle, a reasonable and feasible decision trajectory for the target vehicle on the road can be determined, and the trajectory information of the decision trajectory can be used as the initial trajectory information of the target vehicle in the trajectory planning process.
[0037] S120, constructing a nonlinear proposition for target trajectory planning corresponding to the target vehicle, wherein the nonlinear proposition for target trajectory planning includes: a trajectory cost function and driving constraints, and the driving constraints include: collision avoidance constraints determined based on a tunneling modeling method.
[0038] Among them, the nonlinear proposition of target trajectory planning may refer to a nonlinear programming (NonLinear Programming, NLP) proposition for trajectory planning tasks. The nonlinear proposition of target trajectory planning can be used as an optimal control proposition so that it can be solved using a computational optimal control (Computational Optimal Control) method or a numerical optimal control (Numerical Optimal Control) method. The trajectory planning task may refer to planning a driving trajectory that meets the driving constraints between the motion state of the target vehicle at the starting moment and the motion state at the ending moment. The collision avoidance constraint may refer to a constraint to avoid collision with obstacles in the driving scene. Figure 2 An example of a driving scenario is given. Figure 2As shown, the driving scene may include stationary obstacles and moving obstacles as well as scattered points used to describe the edges of the road, and the collision avoidance constraint condition may be used to describe the condition that the target vehicle (i.e., vehicle i) does not overlap with all obstacles and the edges of the road. The tunneling modeling method may be a method of laying a local tunnel in time and space based on the initial trajectory information to separate the actual drivable area of the target vehicle from the obstacles in the environment. The trajectory cost function may refer to the optimization objective function of the nonlinear proposition of the target trajectory planning, which is an indicator for screening high-quality trajectories. Usually, there are multiple trajectories that meet the driving constraints, so it is necessary to screen out the optimal trajectory as the final result based on the trajectory cost function. Since there are driving constraints and indicators for optimization in the trajectory planning task, it is suitable to use the form of an optimal control problem to describe the trajectory planning proposition of the autonomous driving vehicle.
[0039] Specifically, based on the tunnel modeling method, the tunnel area where the target vehicle can drive at each moment can be determined, thereby converting the complex collision avoidance constraints directly established based on all obstacles in the road into simple collision avoidance constraints in the form of boundary constraints, and only the obstacles around the target vehicle need to be considered, without considering obstacles that are far away and will not affect the driving of the target vehicle, thereby greatly reducing the complexity of the collision avoidance constraints. The corresponding trajectory cost function can be determined based on the expectation that the trajectory is as smooth as possible and as close to the decision trajectory as possible. Based on the trajectory cost function and the collision avoidance constraints, a simpler nonlinear proposition for target trajectory planning can be constructed using the optimal control problem method.
[0040] S130. Solve the nonlinear proposition of target trajectory planning according to the initial trajectory information to determine the target planning trajectory corresponding to the target vehicle.
[0041] Specifically, the initial trajectory information can be used as the initial solution to solve the nonlinear proposition of the target trajectory planning, that is, the trajectory cost function is minimized while satisfying the collision avoidance constraints determined based on the tunneling modeling method, so that the optimal solution can be obtained, that is, the target planning trajectory corresponding to the target vehicle is determined. In this embodiment, the interior point method (IPM) can be used to solve the nonlinear proposition of the target trajectory planning. In this embodiment, by solving a simpler nonlinear proposition of the target trajectory planning, redundant calculations can be avoided, thereby improving the efficiency of trajectory planning.
[0042] The technical solution of this embodiment determines the collision avoidance constraints based on the tunnel modeling method, that is, obtains the drivable tunnel area of the target vehicle at each moment, so that there is no need to consider obstacles that will not affect the driving of the target vehicle during the trajectory planning process, thereby reducing the complexity of the collision avoidance constraints, thereby constructing a simpler target trajectory planning nonlinear proposition, and can quickly solve the target trajectory planning nonlinear proposition by minimizing the trajectory cost function to obtain the target planning trajectory corresponding to the target vehicle, thereby avoiding redundant calculations and improving trajectory planning efficiency.
[0043] On the basis of the above technical solution, the driving constraints may also include: kinematic constraints of the target vehicle, boundary value constraints and internal mechanical constraints of the vehicle.
[0044] The kinematic constraints of the target vehicle may refer to the motion capacity restrictions of the target vehicle during driving. The boundary constraints may refer to the motion state specified by the start and end times of the target vehicle. The internal mechanical constraints of the vehicle may refer to the internal motion capacity restrictions of the target vehicle.
[0045] Exemplarily, a two-degree-of-freedom model may be used to describe vehicle motion. Figure 3 An example of a two-degree-of-freedom vehicle kinematic model is given. Figure 3 As shown in the figure, the two-degree-of-freedom model can merge the two front wheels and the two rear wheels of the vehicle into a virtual single wheel in the longitudinal axis direction of the vehicle body. By determining the rotation angular velocity of the virtual front wheel and the linear acceleration variable of the virtual rear wheel, the front wheel turning angle, driving speed, etc. of the vehicle can be indirectly determined, thereby realizing the vehicle movement.
[0046] based on Figure 3 , the kinematic constraints of the target vehicle can be expressed as:
[0047]
[0048] Among them, the driving time t∈[0,t f ], t f Represents the end time, which can be a constant or a variable; (x i (t),y i (t)) represents the coordinate of the midpoint of the rear wheel axle of vehicle i in the Cartesian coordinate system XOY (i.e., rectangular coordinate system); v i (t) and a i (t) represent the velocity and acceleration along the longitudinal axis of the vehicle, respectively, so that the vehicle's forward direction is the positive direction; φ i (t) represents the front wheel deflection angle of the vehicle, with the left turn direction as the positive direction; ω i(t) represents the angular velocity of the front wheel, with the positive direction perpendicular to the XOY coordinate system and outward; θ i (t) represents the attitude angle of the vehicle in the XOY coordinate system, that is, the rotation angle from the positive direction of the coordinate system's X-axis to the positive direction of the vehicle's longitudinal axis, with counterclockwise rotation as the positive direction. i (t), y i (t),φ i (t),θ i (t) and v i (t) belongs to the state variable x(t), a i (t) and ω i (t) belongs to the control variable u(t). If the motion state x(0) of vehicle i at the initial moment and [0,t f ]u(t) in the time domain, the motion state x(t) in the time domain can be determined one by one by integration, that is, a specific motion trajectory of vehicle i can be obtained.
[0049] Since t = t f The moment is not the end of the entire driving process in a macro sense, so t f Set to the trajectory decision time domain length t decision The preset ratio, t f =t decision γ rate , where 0<γ rate <1 represents the preset proportionality factor.
[0050] based on Figure 3 , the starting point constraint in the boundary value constraint can be expressed as:
[0051] [v i (0),φ i (0),x i (0),y i (0),θ i (0)]=[v 0i ,p 0i ,x 0i ,y 0i ,θ 0i ] (2)
[0052] Among them, [v 0i ,φ 0i ,x 0i ,y 0i ,θ 0i ] can be the motion state information collected by the vehicle-mounted sensor. End time t f It is not the end point of the driving process at a macro level, so there is no need to impose hard restrictions on the end of the local planning time domain.
[0053] based on Figure 3 , the vehicle internal mechanical constraints can be expressed as:
[0054] |φ i (t)|≤Φ max ,|a i (t)|≤a max ,|v i (t)|≤v max ,|ω i (t)|≤Ω max ,t∈[0,t f ]. (3)
[0055] Among them, Φ max 、a max 、v max and Ω max are the maximum amplitudes of each state variable and control variable respectively. Φ max Represents the vehicle's front wheel turning angle φ i (t) The maximum allowable deflection angle value; v max is the upper limit of the vehicle's safe driving speed in low-speed scenarios; to ensure passenger comfort, a max With Ω max are the maximum amplitudes of linear acceleration and front wheel angular velocity, respectively. For example, if the acceleration variable needs to change smoothly, the differential variable jerk of the acceleration can be added i (t) = da i (t) / dt and limit it.
[0056] based on Figure 3 , the trajectory cost function of target vehicle i can be expressed as:
[0057]
[0058] Among them, (x decision (t),y decision (t),θ decision (t)) represents the decision trajectory of target vehicle i.
[0059] Specifically, by taking the collision avoidance constraints determined based on the tunneling modeling method, the kinematic constraints of the target vehicle, the boundary value constraints and the internal mechanical constraints of the vehicle as driving constraints, the trajectory cost function is minimized, so that a more accurate target planning trajectory can be obtained, further improving the accuracy of trajectory planning.
[0060] It should be noted that in this embodiment, the trajectory planning is performed in a Cartesian coordinate system XOY that does not rely on the guide line, so that the trajectory planning task on the road can be completely independent of the guide line.
[0061] On the basis of the above technical scheme, the collision avoidance constraint conditions are determined based on the tunnel modeling method, which may include: covering the body area of the target vehicle with two circular areas of preset radius, and determining the motion trajectories of the two circle centers; determining the local tunnel area corresponding to each circle center at each sampling time in the trajectory planning time domain according to the motion trajectory of the two circle centers and the preset radius; and determining the collision avoidance constraint conditions according to the local tunnel area.
[0062] The preset radius may be set based on the size of the target vehicle. For example, the radius of two circumscribed circles of the target vehicle may be used as the preset radius. Figure 3 As shown, the front and rear wheelbase L of the vehicle can be w 、Vehicle front overhang distance L f 、Vehicle rear overhang distance L r and vehicle width L b Determine the preset radius. For example, the preset radius corresponding to the target vehicle i is: The local tunnel area corresponding to each circle center may refer to a drivable area of the target vehicle.
[0063] Specifically, the target vehicle can be characterized by the centers of the two circumscribed circles of the target vehicle, and a corresponding local tunnel area can be constructed for each center, so as to realize the tunnel modeling of the target vehicle. i The circle evenly covers the rectangular body area of the target vehicle. Based on the initial trajectory information (x decision (t),y decision (t),θ decision (t)) and the corresponding relationship between the coordinates of the center of the circle and the coordinates of the midpoint of the rear wheel axle of the vehicle position, the motion trajectories of the two center of the circle can be determined, which are recorded as: P r =(xr(t),yr(t)) and P f =(xf(t),yf(t)). The corresponding relationship between the coordinates of the center of the circle and the coordinates of the midpoint of the rear wheel axle of the vehicle can be expressed as follows:
[0064]
[0065] It can be seen that the constraint that the vehicle does not collide with the obstacle can be converted to: P r and P f Keep at least a distance R from obstacles i , so that by limiting P r and P f The value range of P r and P f Keep at least a distance R from obstacles i, and then obtain simpler collision avoidance constraints. For example, based on the discretization precision parameter N in numerical optimization fe , for trajectory planning in the time domain [0,t f ] uniform sampling (N fe +1) moments, among which the k The time is t k =t f (k-1) / N fe For each sampling moment, the local tunnel area corresponding to each circle center can be determined, and then P can be obtained based on the local tunnel area. r and P f The value range of is the collision avoidance constraint condition. This embodiment uses two circle centers as mass points to characterize the target vehicle, and constructs a corresponding local tunnel area for each circle center, so that the tunnel modeling process of the target vehicle can be realized more quickly, and a simpler collision avoidance constraint condition can be obtained, thereby further improving the trajectory planning efficiency.
[0066] Exemplarily, determining collision avoidance constraints based on the local tunnel area may include: determining the position coordinates of four area vertices of the local tunnel area in a Cartesian coordinate system; determining the constraint range of the position coordinates of each circle center at each sampling moment based on the position coordinates of the four area vertices, to obtain the collision avoidance constraints.
[0067] The local tunnel area may be a rectangular area. The area vertices may refer to the four vertices of the rectangular area. For example, Figure 4 An example of a local tunnel region is given. Figure 4 Given the center P r The motion trajectory P r (t) and at sampling time t k The center of the circle P r Position P r (t k ).exist Figure 4 The small rectangular area represents the sampling time t k Time center P r The corresponding local tunnel area. The large rectangular area boundary represents the critical point where the target vehicle collides with the obstacle. Figure 4 The distance between the two rectangular areas can be a preset radius R i , to ensure P r Keep at least a distance R from obstacles i .
[0068] Specifically, it can be Figure 3The position coordinates of the four vertices of the local tunnel area are obtained in the Cartesian coordinate system XOY in , and the position coordinates of the corresponding center of the circle can be restricted based on the position coordinates of the four vertices of the area to ensure that the position coordinates of the corresponding center of the circle are located in the local tunnel area, thereby obtaining the collision avoidance constraint condition. For example, Figure 4 The center of the circle is P r =(xr(t),yr(t)) at sampling time t k The constraint range of the center position coordinates at x is: r_lb_k ≤xr(t k )≤x r_ub_k ,y r_lb_k ≤yr(t k )≤y r_ub_k Similarly, we can also use the center P f Corresponding local tunnel area, determine the center P f The constraint range of the position coordinates of . The collision avoidance constraint condition formed by a simple boundary in this embodiment can be expressed as:
[0069]
[0070] It should be noted that no matter how many obstacles there are in the driving scene, there will only be 2 (N fe +1) sets of constraints in the form of (6), thus proposing the stability of the solution.
[0071] For example, the target trajectory planning nonlinear proposition P NLP0 It can be fully expressed as:
[0072] Minimize: Trajectory cost function (4)
[0073] st kinematic constraints (1)
[0074] Boundary value constraints (2)
[0075] Mechanical constraints within the vehicle(3)
[0076] Center of circle relationship (5)
[0077] Collision avoidance constraints (6)
[0078] Specifically, by solving the above nonlinear proposition P of target trajectory planning NLP0 , the trajectory cost function can be minimized while satisfying all constraints, thereby obtaining a more accurate target planning trajectory and improving the trajectory planning efficiency and accuracy.
[0079] Embodiment 2
[0080] Figure 5A trajectory planning method is provided in the second embodiment of the present invention. Based on the above embodiments, this embodiment optimizes "determining the local tunnel area corresponding to each center of the circle at each sampling time in the trajectory planning time domain according to the motion trajectory of the two centers of the circle and the preset radius". The explanations of the terms that are the same or corresponding to the above embodiments are not repeated here.
[0081] See also Figure 5 The trajectory planning method provided in this embodiment specifically includes the following steps:
[0082] S210: Acquire initial trajectory information obtained by performing trajectory decision on the target vehicle.
[0083] S220: Cover the body area of the target vehicle with two circular areas of preset radius, and determine the movement trajectories of the two circle centers.
[0084] Specifically, based on the initial trajectory information (x decision (t),y decision (t),θ decision (t)) and the corresponding relationship between the coordinates of the center of the circle and the coordinates of the midpoint of the rear wheel axle of the vehicle position, the motion trajectories of the two center of the circle can be determined, which are recorded as: P r =(xr(t),yr(t)) and P f =(xf(t),yf(t)).
[0085] S230 . Determine the center positions of the two circle centers at each sampling time in the planning time domain according to the movement trajectories of the two circle centers.
[0086] Specifically, for each circle center, the position of the circle center at each sampling time can be determined based on the movement trajectory of the circle center. r The motion trajectory P r (t), it can be determined that at the sampling time t k The center of the circle P r The center position of the circle is P r (t k ).
[0087] S240: Based on the preset radius, incremental expansion is performed with the center position of each circle as the center to determine the local tunnel area corresponding to each circle center at each sampling time.
[0088] Specifically, with the center position of the circle at each sampling moment as the center, an area range of a preset radius can be obtained, and by incrementally expanding around the area range, the local tunnel area of each center of the circle at each sampling moment can be determined. This embodiment can obtain an accurate local tunnel area by incrementally expanding each center position, thereby improving the accuracy of determining the collision avoidance constraint conditions and further improving the accuracy of trajectory planning.
[0089] Exemplarily, S240 may include: taking the center position of the current circle at the current sampling moment as the center, and extending the distance of the preset radius along the positive and negative directions of the X-axis and the Y-axis respectively, to construct a quadrilateral centered on the center position of the circle; successively expanding outward by the preset distance along the four expansion sides of the quadrilateral to obtain the expansion area of each expansion; if it is detected that the expansion area does not overlap with the obstacle area in the driving scene, updating the quadrilateral based on the expansion area, and continuing the expansion operation on the updated quadrilateral; if it is detected that the expansion area overlaps with the obstacle area in the driving scene, stopping the expansion operation of the expansion side corresponding to the expansion area, and continuing the expansion operation of other expansion sides; when the expansion operations of the four expansion sides are all stopped, determining the local tunnel area corresponding to the current center of the circle at the current sampling moment according to the current quadrilateral.
[0090] Among them, any sampling moment can be used as the current sampling moment, and any circle center can be used as the current circle center, so that the local tunnel area of each circle center at each sampling moment can be determined by using the above-mentioned determination method of the local tunnel area.
[0091] Specifically, Figure 6 An example of the construction process of a local tunnel area is given. Figure 6 As shown in a) in the figure, the current sampling time t k Current circle center P r The center position P r (t k ) as the center, along the positive and negative directions of the X-axis and Y-axis in the XOY coordinate system, the lengths are extended to the preset radius R i distance, construct the circle with the center position P r (t k ) is the center of a quadrilateral, that is, a rectangle. Figure 6 As shown in b) in FIG. 1 , the four expansion sides of the quadrilateral can be expanded outward by a preset distance Δstep in a clockwise or counterclockwise direction to obtain the expansion area of each expansion, such as the area marked by the serial number in FIG. b). The order of the serial numbers can be used to characterize the expansion sequence. After each expansion, it can be detected whether the expansion area overlaps with the obstacle area in the driving scene. For example, the moving obstacle in the driving scene can be detected at t kThe position at the moment, the position of the stationary obstacle, and the scattered point position of the road edge are all stored in the set V, so that by detecting whether there is an element position contained in the expansion area in the set V, it can be determined whether there is an overlapping part. If not, it means that the target vehicle will not collide with the obstacle when driving in the expansion area, that is, the expansion area is a valid area. At this time, the expansion area can be merged into the quadrilateral to update the quadrilateral, and the expansion operation of the updated quadrilateral can be continued. If there is an overlapping part between the expansion area and the obstacle area in the driving scene, it means that the target vehicle will collide with the obstacle when driving in the expansion area, that is, the expansion area is an invalid area. At this time, there is no need to merge the expansion area, stop the expansion operation of the expansion edge corresponding to the expansion area, and continue the expansion operation of other expansion edges. If it is detected that the expansion operations of the four expansion edges have all stopped, it means that the current quadrilateral obtained at the current moment is the current center P r The maximum movable area is shown in Figure c). In order to ensure that the target vehicle does not collide with obstacles, the current quadrilateral area corresponding to the current circle center can be reduced to obtain the local tunnel area corresponding to the current circle center at the current sampling time.
[0092] Exemplarily, determining the local tunnel area corresponding to the current center of the circle at the current sampling time according to the current quadrilateral may include: shrinking the four sides of the current quadrilateral toward the center by a distance of a preset radius, and determining the area of the shrunk quadrilateral as the local tunnel area corresponding to the current center of the circle at the current sampling time. Figure 6 As shown in d), by shrinking the four sides of the current quadrilateral to the center to a preset radius R i The distance between each circle center and the obstacle is at least R. i , avoiding the collision between the target vehicle and the obstacle.
[0093] S250: Determine collision avoidance constraint conditions according to the local tunnel area.
[0094] S260: construct a nonlinear proposition for target trajectory planning corresponding to the target vehicle, wherein the nonlinear proposition for target trajectory planning includes: a trajectory cost function and driving constraints, and the driving constraints include: a collision avoidance constraint determined based on a tunneling modeling method.
[0095] S270: Solve the nonlinear proposition of target trajectory planning according to the initial trajectory information to determine the target planning trajectory corresponding to the target vehicle.
[0096] The technical solution of this embodiment can accurately obtain the local tunnel area by incrementally expanding each center position of the circle at each sampling moment, thereby improving the accuracy of determining the collision avoidance constraint conditions, and further improving the accuracy of trajectory planning.
[0097] Embodiment 3
[0098] Figure 7 A trajectory planning method is provided in the third embodiment of the present invention. Based on the above embodiments, this embodiment optimizes "constructing a nonlinear proposition of target trajectory planning corresponding to the target vehicle", and on this basis, further optimizes "solving the nonlinear proposition of target trajectory planning according to the initial trajectory information to determine the target planning trajectory corresponding to the target vehicle". The explanations of the terms that are the same or corresponding to the above embodiments are not repeated here.
[0099] See also Figure 7 The trajectory planning method provided in this embodiment specifically includes the following steps:
[0100] S310: Acquire initial trajectory information obtained by performing trajectory decision on the target vehicle.
[0101] S320. Construct a first trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the first trajectory planning nonlinear proposition includes: an objective function and boundary constraints, wherein the objective function is determined based on a trajectory cost function, a kinematic constraint of the target vehicle, and a center correlation relationship between two circle centers corresponding to a collision avoidance constraint determined based on a tunneling modeling method.
[0102] The boundary constraint condition may refer to a condition that is restricted by a boundary form. For example, the boundary constraint condition may include a collision avoidance constraint condition determined based on a tunnel modeling method, a boundary value constraint condition, and a vehicle internal mechanical constraint condition. The objective function may refer to the first trajectory planning nonlinear proposition P NLP1 optimization goal.
[0103] Specifically, due to the nonlinear proposition P NLP0 The two centers P r and P f , and the optimization variable x i ,y i and θ i A nonlinear relationship is formed (see equation (5)), so that the collision avoidance constraint (see equation (6)) is a non-convex constraint. It can be seen that the nonlinear proposition P of target trajectory planning NLP0 The non-convex nonlinear constraints in can include the kinematic constraints of the target vehicle (see equation (1)) and the center-to-center relationship of the two circles (see equation (5)), which leads to the nonlinear proposition P for target trajectory planning:NLP0 The solution efficiency is low. To solve this problem, xr(t), yr(t), xf(t) and yf(t) can also be optimized as optimization variables, that is, the first trajectory planning nonlinear proposition P is constructed. NLP1 The optimization variables of the objective function in include: xr(t), yr(t), xf(t), yf(t), x i ,y i and θ i , so that the first trajectory planning nonlinear proposition P NLP1 The constraints in are all linear boundary constraints.
[0104] Exemplarily, determining the objective function based on the trajectory cost function, the kinematic constraints of the target vehicle, and the center correlation relationship of two circle centers corresponding to the collision avoidance constraints determined based on the tunneling modeling method can include: performing polynomial characterization on the kinematic constraints of the target vehicle and the center correlation relationship of two circle centers corresponding to the collision avoidance constraints determined based on the tunneling modeling method to determine each penalty term; and determining the objective function based on each penalty term and the trajectory cost function.
[0105] For example, the kinematic constraints of the target vehicle (see formula (1)) and the center-to-center relationship of the two circles (see formula (5)) can be transformed into penalty polynomials to obtain the corresponding penalty terms, and the trajectory cost function and the penalty terms are added to obtain the objective function. For example, the objective function can be expressed as follows:
[0106]
[0107] For example, the first trajectory planning nonlinear proposition P NLP1 It can be fully expressed as:
[0108] Minimize: Objective function (7)
[0109] st boundary value constraint condition (2)
[0110] Mechanical constraints within the vehicle(3)
[0111] Collision avoidance constraints (6)
[0112] S330: Solve the nonlinear proposition of the first trajectory planning according to the initial trajectory information to determine the first planned trajectory corresponding to the target vehicle.
[0113] Specifically, the initial trajectory information is used as the initial solution to plan the nonlinear proposition P for the first trajectory. NLP1 By solving, we can obtain the first planning trajectory corresponding to the target vehicle. NLP1Temporarily ignore the trajectory cost function (4) and focus on obtaining a more kinematically consistent first planning trajectory while ensuring no collision. NLP1 The constraints are all simple boundary constraints, so P NLP1 There must be a solution that allows using IPM to solve P NLP1 The solution speed is higher than P NLP0 The solution speed.
[0114] It should be noted that this embodiment does not limit the execution order of S330 , and step S330 may be executed after step S320 or after step S340 .
[0115] S340: construct a second trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the second trajectory planning nonlinear proposition includes: a trajectory cost function and driving constraints.
[0116] Specifically, the second trajectory planning nonlinear proposition P NLP2 The nonlinear proposition P NLP0 The optimization objectives of are consistent, both are trajectory cost functions. Driving constraints can include collision avoidance constraints determined based on tunnel modeling, and can also include: kinematic constraints of the target vehicle, boundary constraints, and internal mechanical constraints of the vehicle.
[0117] For example, the second trajectory planning nonlinear proposition P NLP2 It can be fully expressed as:
[0118] Minimize: Trajectory cost function (4)
[0119] st kinematic constraints (1)
[0120] Boundary value constraints (2)
[0121] Mechanical constraints within the vehicle(3)
[0122] Center of circle relationship (5)
[0123] Collision avoidance constraints (6)
[0124] For example, based on the tunnel modeling method provided in the above embodiments, the nonlinear proposition P of the second trajectory planning can be re-determined according to the trajectory information of the first planning trajectory and the center correlation relationship of the two circle centers (see formula (5)). NLP2 The collision avoidance constraints in (6) are added to further improve the accuracy of path planning.
[0125] S350: Solve the nonlinear proposition of the second trajectory planning according to the first planned trajectory to determine the target planned trajectory corresponding to the target vehicle.
[0126] Specifically, the first trajectory information is used as the initial solution to plan the nonlinear proposition P for the second trajectory. NLP2 By solving, we can obtain the target planning trajectory corresponding to the target vehicle. NLP2 Under the premise of ensuring kinematic feasibility, we focus on minimizing the trajectory cost function (5), so that the target planning trajectory can be solved more quickly. Since there are three tasks that must be completed simultaneously in the nonlinear proposition of target trajectory planning, namely, hard vehicle kinematic constraints, hard collision avoidance constraints, and trajectory cost function optimization, the nonlinear proposition of target trajectory planning is difficult to solve. By decomposing the nonlinear proposition of target trajectory planning into the first trajectory planning nonlinear proposition and the second trajectory planning nonlinear proposition, the difficulty of solving can be reduced and the efficiency of trajectory planning can be further improved.
[0127] The technical solution of this embodiment can reduce the difficulty of solving the trajectory planning task and further improve the trajectory planning efficiency by solving the first trajectory planning nonlinear proposition and the second trajectory planning nonlinear proposition step by step.
[0128] The following is an embodiment of a trajectory planning device provided in an embodiment of the present invention. The device and the trajectory planning methods of the above-mentioned embodiments belong to the same inventive concept. For details not described in detail in the embodiment of the trajectory planning device, reference can be made to the embodiment of the above-mentioned trajectory planning method.
[0129] Embodiment 4
[0130] Figure 8 This is a schematic diagram of the structure of a trajectory planning device provided in the fourth embodiment of the present invention. This embodiment is applicable to the case of trajectory planning for an autonomous driving vehicle, and is particularly applicable to the case of trajectory planning for an autonomous driving vehicle on a structured road. Figure 8 As shown, the device specifically includes: an initial trajectory information acquisition module 410, a trajectory planning nonlinear proposition construction module 420 and a target planning trajectory determination module 430.
[0131] Among them, the initial trajectory information acquisition module 410 is used to obtain the initial trajectory information obtained by making a trajectory decision for the target vehicle; the trajectory planning nonlinear proposition construction module 420 is used to construct a target trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the target trajectory planning nonlinear proposition includes: a trajectory cost function and driving constraints, and the driving constraints include: collision avoidance constraints determined based on the tunneling modeling method; the target planning trajectory determination module 430 is used to solve the target trajectory planning nonlinear proposition according to the initial trajectory information to determine the target planning trajectory corresponding to the target vehicle.
[0132] The technical solution of this embodiment determines the collision avoidance constraints based on the tunnel modeling method, that is, obtains the drivable tunnel area of the target vehicle at each moment, so that there is no need to consider obstacles that will not affect the driving of the target vehicle during the trajectory planning process, thereby reducing the complexity of the collision avoidance constraints, thereby constructing a simpler target trajectory planning nonlinear proposition, and can quickly solve the target trajectory planning nonlinear proposition by minimizing the trajectory cost function to obtain the target planning trajectory corresponding to the target vehicle, thereby avoiding redundant calculations and improving trajectory planning efficiency.
[0133] Optionally, the device further includes: a collision avoidance constraint condition determination module, including:
[0134] A motion trajectory determination submodule is used to cover the body area of the target vehicle with two circular areas of preset radius and determine the motion trajectories of the two circle centers;
[0135] A local tunnel area determination submodule is used to determine the local tunnel area corresponding to each circle center at each sampling time in the trajectory planning time domain according to the motion trajectories of the two circle centers and the preset radius;
[0136] The collision avoidance constraint condition determination submodule is used to determine the collision avoidance constraint condition according to the local tunnel area.
[0137] Optionally, the local tunnel area determination submodule includes:
[0138] A circle center position determination unit, used to determine the circle center positions of the two circle centers at each sampling time in the planning time domain according to the motion trajectories of the two circle centers;
[0139] The local tunnel area determination unit is used to perform incremental expansion based on a preset radius and with the center position of each circle as the center to determine the local tunnel area corresponding to each circle center at each sampling time.
[0140] Optionally, the local tunnel area determination unit is specifically configured to:
[0141] Taking the center position of the current circle at the current sampling moment as the center, the distance of the preset radius is extended along the positive and negative directions of the X-axis and the Y-axis respectively to construct a quadrilateral with the center position of the circle as the center; the preset distance is successively expanded outward along the four expansion edges of the quadrilateral to obtain the expansion area of each expansion; if it is detected that the expansion area does not overlap with the obstacle area in the driving scene, the quadrilateral is updated based on the expansion area, and the expansion operation of the updated quadrilateral is continued; if it is detected that the expansion area overlaps with the obstacle area in the driving scene, the expansion operation of the expansion edge corresponding to the expansion area is stopped, and the expansion operation of other expansion edges is continued; when the expansion operations of the four expansion edges are all stopped, the local tunnel area corresponding to the current center of the circle at the current sampling moment is determined according to the current quadrilateral.
[0142] Optionally, the local tunnel area determination unit is further specifically used to: shrink the four sides of the current quadrilateral toward the center by a distance of a preset radius, and determine the area of the shrunk quadrilateral as the local tunnel area corresponding to the current center of the circle at the current sampling moment.
[0143] Optionally, the collision avoidance constraint condition determination submodule is specifically used for:
[0144] The position coordinates of four regional vertices of the local tunnel region in the Cartesian coordinate system are determined; based on the position coordinates of the four regional vertices, the constraint range of the position coordinates of each circle center at each sampling time is determined to obtain the collision avoidance constraint conditions.
[0145] Optionally, the driving constraints also include: kinematic constraints of the target vehicle, boundary value constraints and internal mechanical constraints of the vehicle.
[0146] Optionally, the trajectory planning nonlinear proposition construction module 420 is specifically used to: construct a first trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the first trajectory planning nonlinear proposition includes: an objective function and boundary constraints, and the objective function is determined based on the trajectory cost function, the kinematic constraints of the target vehicle, and the center correlation relationship of two circle centers corresponding to the collision avoidance constraints determined based on the tunneling modeling method; construct a second trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the second trajectory planning nonlinear proposition includes: a trajectory cost function and a driving constraint;
[0147] Accordingly, the target planning trajectory determination module 430 is specifically used to: solve the first trajectory planning nonlinear proposition according to the initial trajectory information to determine the first planning trajectory corresponding to the target vehicle; solve the second trajectory planning nonlinear proposition according to the first planning trajectory to determine the target planning trajectory corresponding to the target vehicle.
[0148] Optionally, the device also includes an objective function determination module, which is used to: perform polynomial representation on the center correlation relationship between the two circle centers corresponding to the kinematic constraints of the target vehicle and the collision avoidance constraints determined based on the tunneling modeling method, and determine each penalty term; determine the objective function based on each penalty term and the trajectory cost function.
[0149] The trajectory planning device provided in the embodiment of the present invention can execute the trajectory planning method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the trajectory planning method.
[0150] It is worth noting that in the embodiment of the above-mentioned trajectory planning device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.
[0151] Embodiment 5
[0152] Fig. 9 A schematic diagram of the structure of a device provided in Embodiment 5 of the present invention. Fig. 9 A block diagram of an exemplary device 12 suitable for use in implementing embodiments of the present invention is shown. Fig. 9 The device 12 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.
[0153] like Fig. 9 As shown, device 12 is in the form of a general purpose computing device. Components of device 12 may include, but are not limited to, one or more processors or processing units 16, system memory 28, and bus 18 that connects various system components (including system memory 28 and processing unit 16).
[0154] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.
[0155] Device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by device 12, including volatile and non-volatile media, removable and non-removable media.
[0156] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Fig. 9 not shown, usually called a "hard drive"). Although Fig. 9 Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.
[0157] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. Program modules 42 generally perform the functions and / or methods of the embodiments described herein.
[0158] The device 12 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), may communicate with one or more devices that enable a user to interact with the device 12, and / or may communicate with any device that enables the device 12 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the device 12 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with other modules of the device 12 via a bus 18. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the device 12, including, but not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0159] The processing unit 16 executes various functional applications and data processing by running the program stored in the system memory 28, for example, implementing a trajectory planning method provided by the embodiment of the present invention, the method comprising:
[0160] Obtaining initial trajectory information obtained by making trajectory decisions for the target vehicle;
[0161] Constructing a nonlinear proposition of target trajectory planning corresponding to the target vehicle, wherein the nonlinear proposition of target trajectory planning includes: trajectory cost function and driving constraints, and the driving constraints include: collision avoidance constraints determined based on tunneling modeling;
[0162] According to the initial trajectory information, the nonlinear proposition of target trajectory planning is solved to determine the target planning trajectory corresponding to the target vehicle.
[0163] Of course, those skilled in the art can understand that the processor can also implement the technical solution of the trajectory planning method provided by any embodiment of the present invention.
[0164] Embodiment 5
[0165] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps of the trajectory planning method according to any embodiment of the present invention are implemented. The method includes:
[0166] Obtaining initial trajectory information obtained by making trajectory decisions for the target vehicle;
[0167] Constructing a nonlinear proposition of target trajectory planning corresponding to the target vehicle, wherein the nonlinear proposition of target trajectory planning includes: trajectory cost function and driving constraints, and the driving constraints include: collision avoidance constraints determined based on tunneling modeling;
[0168] According to the initial trajectory information, the nonlinear proposition of target trajectory planning is solved to determine the target planning trajectory corresponding to the target vehicle.
[0169] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.
[0170] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.
[0171] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
[0172] Computer program code for performing the operations of the present invention may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0173] It should be understood by those skilled in the art that the modules or steps of the present invention described above can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, optionally, they can be implemented by a program code executable by a computer device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.
[0174] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims
1. A trajectory planning method, characterized in that: include: Obtaining initial trajectory information obtained by making trajectory decisions for the target vehicle; Constructing a nonlinear proposition of target trajectory planning corresponding to the target vehicle, wherein the nonlinear proposition of target trajectory planning includes: a trajectory cost function and driving constraints, wherein the trajectory cost function is determined based on the trajectory being as smooth as possible and as close as possible to the decision trajectory; the driving constraints include: collision avoidance constraints determined based on a tunneling modeling method; the tunneling modeling method is a method of laying a local tunnel in time and space based on the initial trajectory information to separate the actual drivable area of the target vehicle from obstacles in the environment; The target trajectory planning nonlinear proposition is solved according to the initial trajectory information to determine the target planning trajectory corresponding to the target vehicle.
2. The method according to claim 1, characterized in that The collision avoidance constraints are determined based on the tunnel modeling approach, including: Covering the body area of the target vehicle with two circular areas of preset radius, and determining the motion trajectories of the two circle centers; Determine the local tunnel area corresponding to each circle center at each sampling time in the trajectory planning time domain according to the motion trajectories of the two circle centers and the preset radius; A collision avoidance constraint condition is determined according to the local tunnel area.
3. The method according to claim 2, characterized in that According to the motion trajectories of the two circle centers and the preset radius, determining the local tunnel area corresponding to each circle center at each sampling time in the trajectory planning time domain includes: According to the motion trajectories of the two circle centers, the center positions of the two circle centers at each sampling time in the planning time domain are determined; Based on the preset radius, incremental expansion is performed with the center position of each circle as the center to determine the local tunnel area corresponding to each circle center at each sampling time.
4. The method according to claim 3, characterized in that: Based on the preset radius, incremental expansion is performed with the center position of each circle as the center to determine the local tunnel area corresponding to each circle center at each sampling time, including: Taking the center position of the current circle at the current sampling moment as the center, extending along the positive and negative directions of the X-axis and the Y-axis by a distance of a preset radius respectively, to construct a quadrilateral centered on the center position of the circle; Expand outwards successively along the four expansion sides of the quadrilateral by a preset distance respectively, to obtain an expansion area for each expansion; If it is detected that there is no overlapping portion between the expanded area and the obstacle area in the driving scene, updating the quadrilateral based on the expanded area, and continuing the expansion operation on the updated quadrilateral; If it is detected that the expansion area overlaps with the obstacle area in the driving scene, the expansion operation of the expansion edge corresponding to the expansion area is stopped, and the expansion operations of other expansion edges are continued; When the expansion operations of the four expansion edges are stopped, the four edges of the current quadrilateral are shrunk toward the center by a distance of a preset radius, and the area of the shrunk quadrilateral is determined as the local tunnel area corresponding to the current circle center at the current sampling moment.
5. The method according to claim 2, characterized in that: According to the local tunnel area, collision avoidance constraint conditions are determined, including: Determine the position coordinates of four area vertices of the local tunnel area in a Cartesian coordinate system; According to the position coordinates of the vertices of the four regions, the constraint range of the position coordinates of each circle center at each sampling moment is determined to obtain the collision avoidance constraint conditions.
6. The method according to claim 1, characterized in that The driving constraints also include: kinematic constraints of the target vehicle, boundary value constraints and internal mechanical constraints of the vehicle.
7. The method according to any one of claims 1 to 6, characterized in that: Constructing a nonlinear proposition for target trajectory planning corresponding to the target vehicle, including: Constructing a first trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the first trajectory planning nonlinear proposition includes: an objective function and boundary constraints, wherein the objective function is determined based on the trajectory cost function, the kinematic constraints of the target vehicle, and the center correlation relationship of two circle centers corresponding to the collision avoidance constraints determined based on the tunneling modeling method; the center correlation relationship of the two circle centers is the correspondence between the center position coordinates and the coordinates of the midpoint of the rear wheel axle of the vehicle position; the objective function is the optimization target of the first trajectory planning nonlinear proposition; the boundary constraints include the collision avoidance constraints determined based on the tunneling modeling method, the boundary value constraints, and the vehicle internal mechanical constraints; Constructing a second trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the second trajectory planning nonlinear proposition includes: a trajectory cost function and a driving constraint condition; the optimization target of the second trajectory planning nonlinear proposition is consistent with that of the target trajectory planning nonlinear proposition, both of which are trajectory cost functions; the driving constraint condition includes a collision avoidance constraint condition determined based on a tunneling modeling method, a kinematic constraint condition of the target vehicle, a boundary value constraint condition, and a vehicle internal mechanical constraint condition; Wherein, the first trajectory planning nonlinear proposition and the second trajectory planning nonlinear proposition are obtained by decomposing the target trajectory planning nonlinear proposition; Accordingly, according to the initial trajectory information, the nonlinear proposition of the target trajectory planning is solved to determine the target planning trajectory corresponding to the target vehicle, including: Solving the first trajectory planning nonlinear proposition according to the initial trajectory information to determine the first planned trajectory corresponding to the target vehicle; According to the first planned trajectory, the nonlinear proposition of the second trajectory planning is solved to determine the target planned trajectory corresponding to the target vehicle.
8. The method according to claim 7, characterized in that The objective function is determined based on the trajectory cost function, the kinematic constraints of the target vehicle, and the center correlation relationship between the two circle centers corresponding to the collision avoidance constraints determined based on the tunneling modeling method, including: The center correlation relationship between the two circle centers corresponding to the kinematic constraints of the target vehicle and the collision avoidance constraints determined based on the tunneling modeling method is characterized by a polynomial, and each penalty term is determined; The objective function is determined according to the various penalty terms and the trajectory cost function.
9. A trajectory planning device, characterized in that: include: An initial trajectory information acquisition module is used to obtain initial trajectory information obtained by making trajectory decisions on the target vehicle; A trajectory planning nonlinear proposition construction module is used to construct a target trajectory planning nonlinear proposition corresponding to the target vehicle, wherein the target trajectory planning nonlinear proposition includes: a trajectory cost function and driving constraints, the trajectory cost function is determined based on the trajectory being as smooth as possible and as close to the decision trajectory as possible; the driving constraints include: collision avoidance constraints determined based on a tunneling modeling method; the tunneling modeling method is a method of laying a local tunnel in time and space based on the initial trajectory information to separate the actual drivable area of the target vehicle from obstacles in the environment; The target planning trajectory determination module is used to solve the target trajectory planning nonlinear proposition according to the initial trajectory information to determine the target planning trajectory corresponding to the target vehicle.
10. A device, characterized in that: The device comprises: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the trajectory planning method as described in any one of claims 1-8.
11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the trajectory planning method as described in any one of claims 1 to 8 is implemented.
