Vehicle trajectory optimization method and device, equipment, storage medium and product
By building the initial grid map and searching the path based on the vehicle collision detection model, the vehicle trajectory is optimized, and the reliability problem of trajectory planning in complex scenarios is solved, the traffic capacity of the autonomous driving system is improved, and the high-precision map and manual labeling are reduced. Dependency of paths.
Patent Information
- Application Number
- CN202411236152.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2024-11-29
AI Technical Summary
The existing trajectory planning technology is difficult to successfully plan a feasible path in complex scenarios, which makes it difficult for the pilot assist system to pass through complex scenarios without the driver taking over. It relies on high-precision maps and manual path labeling, which consumes a lot of manpower and does not reliable.
By building an initial raster map, resolution conversion, binary conversion, and path search based on vehicle collision detection model, optimize vehicle trajectory and reduce dependence on high-precision maps and manual annotation paths.
It improves the reliability of vehicle trajectory optimization in complex scenarios, enhances the passport capability of pilot assisted driving functions, and reduces dependence on high-precision maps and perceived information.
Smart Images

Figure CN119043365A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of path optimization, and in particular to a vehicle trajectory optimization method, device, equipment, storage medium and product. Background Art
[0002] With the development of intelligent driving technology, vehicles equipped with intelligent navigation assistance functions have realized navigation assistance functions on highways, effectively alleviating driving fatigue. However, due to factors such as complex traffic signs and intricate or missing road markings in scenes such as highway toll stations or urban intersections, existing trajectory planning technologies are difficult to successfully plan a feasible path to guide vehicles through complex scenes, resulting in existing navigation assistance systems being difficult to pass through complex scenes without the driver taking over.
[0003] Most of the existing trajectory optimization methods require reference paths provided by high-precision maps or rely on manually annotating drivable paths in complex scenes, and then learn reference paths through complex scenes through deep learning methods. Finally, the autonomous driving system plans a suitable driving path through the complex scenes through the reference path. This method usually relies on manually annotated paths, which consumes a lot of manpower for path data annotation, and it is difficult to obtain reliable reference paths in complex and ever-changing scenes. Summary of the invention
[0004] The present invention provides a vehicle trajectory optimization method, device, equipment, storage medium and product to reduce the dependence on high-precision maps and manually annotated reference path information and improve the reliability of vehicle trajectory optimization in complex scenarios.
[0005] According to one aspect of the present invention, a vehicle trajectory optimization method is provided, the method comprising:
[0006] When the vehicle meets the conditions for entering a complex scene, an initial grid map is constructed based on the vehicle's position information and the obstacle information around the vehicle.
[0007] According to each grid position in the initial grid map and the vehicle grid position of the own vehicle in the initial grid map, the initial grid map is converted into a transformed grid map;
[0008] Binarizing the transformed raster map to obtain a binary raster map;
[0009] Constructing a vehicle collision detection model of the own vehicle, and performing a path search in the binary grid map based on the vehicle collision detection model to obtain a driving path reference line;
[0010] Based on the driving path reference line, the vehicle trajectory of the own vehicle is optimized to obtain a vehicle trajectory optimization result.
[0011] According to another aspect of the present invention, a vehicle trajectory optimization device is provided, the device comprising:
[0012] A map construction module is used to construct an initial grid map based on the vehicle's position information and the obstacle information around the vehicle when the vehicle meets the judgment conditions for entering a complex scene;
[0013] A resolution conversion module, used for performing resolution conversion on the initial grid map according to each grid position in the initial grid map and a vehicle grid position of the own vehicle in the initial grid map to obtain a transformed grid map;
[0014] A binary conversion module, used for performing a binary conversion on the transformed raster map to obtain a binary raster map;
[0015] A model building module is used to build a vehicle collision detection model of the own vehicle, and perform a path search in the binary grid map based on the vehicle collision detection model to obtain a driving path reference line;
[0016] The trajectory optimization module is used to optimize the vehicle trajectory of the own vehicle based on the driving path reference line to obtain a vehicle trajectory optimization result.
[0017] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:
[0018] at least one processor; and
[0019] a memory communicatively connected to the at least one processor; wherein,
[0020] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the vehicle trajectory optimization method described in any embodiment of the present invention.
[0021] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the vehicle trajectory optimization method described in any embodiment of the present invention when executed.
[0022] According to another aspect of the present invention, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the vehicle trajectory optimization method according to any embodiment of the present invention is implemented.
[0023] The above technical solution reduces the dependence on reference paths provided by high-precision maps and perception information, reduces the dependence on manually labeled reference path information, enhances the navigation assisted driving function's traffic capability in complex scenarios, and improves the reliability of vehicle trajectory optimization in complex scenarios.
[0024] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0026] Figure 1A is a flow chart of a vehicle trajectory optimization method provided according to Embodiment 1 of the present invention;
[0027] Figure 1B is a schematic diagram of a method for constructing a vehicle collision detection model provided according to Embodiment 1 of the present invention;
[0028] Figure 2A is a flow chart of a vehicle trajectory optimization method provided according to Embodiment 2 of the present invention;
[0029] Figure 2B is a schematic diagram of a process for determining a grid to be updated according to Embodiment 2 of the present invention;
[0030] Figure 3 is a structural schematic diagram of a vehicle trajectory optimization device provided according to Embodiment 3 of the present invention;
[0031] Figure 4 It is a structural schematic diagram of an electronic device for implementing the vehicle trajectory optimization method according to an embodiment of the present invention. DETAILED DESCRIPTION
[0032] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0033] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0034] Embodiment 1
[0035] Figure 1A This is a flow chart of a vehicle trajectory optimization method provided in the first embodiment of the present invention. This embodiment is applicable to the case of optimizing the trajectory of a vehicle entering a complex scene. The method can be executed by a vehicle trajectory optimization device. The vehicle trajectory optimization device can be implemented in the form of hardware and / or software. The vehicle trajectory optimization device can be configured in an electronic device. Figure 1A As shown, the method includes:
[0036] S110: When it is detected that the vehicle meets the conditions for entering a complex scene, an initial grid map is constructed according to the location information of the vehicle and the obstacle information around the vehicle.
[0037] S120 , performing resolution conversion on the initial grid map according to each grid position in the initial grid map and the vehicle grid position of the own vehicle in the initial grid map to obtain a transformed grid map.
[0038] S130 , binarizing the transformed raster map to obtain a binarized raster map.
[0039] S140: construct a vehicle collision detection model for the own vehicle, and perform a path search in a binary grid map based on the vehicle collision detection model to obtain a driving path reference line.
[0040] S150: Optimize the vehicle trajectory of the own vehicle based on the driving path reference line to obtain a vehicle trajectory optimization result.
[0041] The complex scene may specifically be scenes such as entering a toll booth and passing a fork in the road. The method for determining whether the vehicle has entered a complex scene may be: performing complex scene distance detection based on a navigation map, or detecting whether complex scene feature identification elements appear in front through the vehicle's visual perception module.
[0042] In an optional embodiment, complex scene distance information sent by a map navigation application is obtained. If the complex scene distance information is less than a preset scene distance threshold, it is determined that the vehicle itself meets the judgment conditions for entering a complex scene; or, when a complex scene feature identification element is monitored, it is determined that the vehicle itself meets the judgment conditions for entering a complex scene.
[0043] The map navigation application can be a high-precision map navigation program of a built-in vehicle system or a high-precision map navigation program of an external device, and the external device can establish a communication connection with the vehicle. The scene distance threshold can be pre-set by relevant technical personnel, for example, the scene distance threshold can be set to 200 meters.
[0044] Among them, complex scene feature identification elements may include toll station signs, ETC (Electronic Toll Collection) lane signs, toll sign poles or intersection signs and traffic light signs.
[0045] Exemplarily, the complex scene distance information of the road ahead of the own vehicle is obtained through the map navigation application. If the complex scene distance information is less than the preset scene distance threshold, it is determined that the own vehicle meets the judgment conditions for entering the complex scene; if the complex scene distance information is not less than the preset scene distance threshold, it is determined that the own vehicle does not meet the judgment conditions for entering the complex scene.
[0046] Furthermore, the method for judging whether to enter a complex scene may also be that when a complex scene characteristic identification element is monitored, it is determined that the vehicle itself meets the judgment conditions for entering a complex scene; when no complex scene characteristic identification element is monitored, it is determined that the vehicle itself does not meet the judgment conditions for entering a complex scene.
[0047] The above-mentioned method uses navigation information and environmental perception information as input to plan the travel trajectory in complex scenarios without reference lines such as toll stations or intersections. It does not rely on the initial reference trajectory information provided by high-precision maps, has high flexibility and environmental adaptability, and improves the passing ability of the autonomous driving system in complex scenarios.
[0048] Among them, the vehicle's own position information and the obstacle information around the vehicle are both upstream information of the vehicle. In order to improve the accuracy of the initial grid map construction, the initial grid map can also be constructed based on more upstream information of the vehicle. For example, the upstream information of the vehicle can also include road boundary information, target channel information, surrounding static obstacle information and surrounding traffic participating vehicle information obtained by the upstream sensor of the pilot assisted driving system. The positioning module provides the positioning information of the own vehicle, specifically including the coordinate information, heading angle information, speed and acceleration information of the own vehicle. The surrounding environment and obstacle vehicle information provided by the perception module specifically includes road boundary information, static obstacle boundary information such as cone barrels, and coordinate information, heading angle information, speed and acceleration information of the obstacle vehicle.
[0049] By converting the upstream information of the vehicle into the same coordinate system, usually taking the current position of the own vehicle as the reference point, the upstream information of different vehicles is fused to generate dense point cloud data or image data of the environment; the environmental data is mapped to the initial grid map, the environment is divided into grid cells, and a value is assigned to each grid cell to indicate the occupancy of the area, such as occupied or idle, so as to complete the construction of the initial grid map. The initial grid map can reflect the changes in the environment around the vehicle in real time, which is helpful for navigation and obstacle avoidance.
[0050] It should be noted that in the process of constructing the initial grid map, in order to save computing power and speed up the search speed of the grid, different resolutions can be set for different areas of the initial grid map according to the distance from the vehicle. Specifically, the resolution of the initial grid map can be determined according to the longitudinal distance block, with a higher grid resolution closer to the vehicle and a lower initial grid resolution farther from the vehicle.
[0051] In an optional embodiment, the initial grid map is converted in resolution according to each grid position in the initial grid map and the vehicle grid position of the own vehicle in the initial grid map to obtain a transformed grid map, including: adjusting the resolution corresponding to a first grid area in the initial grid map whose grid position and the vehicle grid position meet a first judgment condition to a first resolution; and adjusting the resolution corresponding to a second grid area in the initial grid map whose grid position and the vehicle grid position meet a second judgment condition to a second resolution; and adjusting the resolution corresponding to a third grid area in the initial grid map whose grid position and the vehicle grid position meet a third judgment condition to a third resolution; generating a transformed grid map including a first grid area of the first resolution, a second grid area of the second resolution, and a third grid area of the third resolution.
[0052] Among them, the grid resolution size relationship between the vehicle grid position and the longitudinal distance grid block in the initial grid map can be as follows:
[0053]
[0054] Where s is the longitudinal distance between the vehicle grid position in the vehicle's current driving direction and the grid position in the initial grid map. 0.2, 0.5, and 1.0 represent the resolution size, respectively.
[0055] Continuing with the above example, the first judgment condition may be that the distance between the grid position and the vehicle grid position satisfies a preset distance threshold of 50 or less; the first grid area is the area where the first grid that satisfies the first judgment condition is located; and the first resolution is 0.2. The second judgment condition may be that the distance between the grid position and the vehicle grid position satisfies a preset distance threshold of 100 or less and a preset distance threshold of 50 or greater; the second grid area is the area where the second grid that satisfies the second judgment condition is located; and the second resolution is 0.5. The third judgment condition may be that the distance between the grid position and the vehicle grid position is greater than or equal to the preset distance threshold of 100; the third grid area is the area where the third grid that satisfies the third judgment condition is located; and the first resolution is 1.0.
[0056] The obtained transformed grid map is a grid map with grid areas of different resolutions. In order to ensure the safety of the planning results, the discrete transformed grid map is subjected to distance transformation, and the distance from each grid in the transformed grid map to the nearest impassable area is calculated. That is, each grid in the converted distance grid map has its corresponding distance value. The grid value of the grid where the impassable area is located can be 0, and the grid values of other grid areas are the distance values between the nearest impassable grid area. The obtained distance grid map is binarized to obtain a binary grid map.
[0057] Based on the preset path search algorithm, the binary grid map is searched for paths. During the search process, the vehicle geometric center based on the constructed vehicle collision detection model can be used as the vehicle position reference point. In order to improve the efficiency of vehicle collision detection, a circular combination can be used to identify the vehicle collision model. Figure 1B A schematic diagram of a method for constructing a vehicle collision detection model is shown. A distance grid map or a binary transformation map can be combined to quickly determine whether the grid currently occupied by the vehicle collides with the surrounding environment. The path search algorithm can be an A* algorithm.
[0058] in, Figure 1B The parameters in are determined as follows:
[0059]
[0060] Wherein, L is the length of the vehicle and W is the width of the vehicle.
[0061] Based on the vehicle collision detection model, a path search is performed in the binary grid map to obtain a driving path reference line.
[0062] For example, a vehicle collision detection search can be performed based on the A* algorithm. In the collision detection process using the A* search, for the extended grid points where the vehicle's heading angle changes, the heading angle corresponding to the collision detection vehicle can be calculated using the following formula:
[0063]
[0064] Among them, x new ,y new Indicates the position coordinates of the newly expanded grid points, x parent ,y parent Indicates the position coordinates of the parent grid point of the newly expanded grid point.
[0065] Use the A* search algorithm to search for paths on the distance grid map or the binary grid map. During the search, the grid search resolution used is determined according to the aforementioned grid resolution determination method for different longitudinal distances, and segmented search is performed. The search results are saved as:
[0066]
[0067] Among them, n coarse Represents the number of discrete grid points of the A* search trajectory result, They represent the horizontal and vertical coordinates of the i-th point in the A* search result respectively.
[0068] The result after A* search only contains the horizontal and vertical coordinate information of the vehicle trajectory, and cannot consider the non-complete constraints of the vehicle. The planning result needs to be smoothed. In the initial smoothing, the A* search planning result is first downsampled at a fixed distance D, and the rough trajectory planned by the A* search method is preliminarily smoothed using the cubic spline curve fitting method. The preliminary trajectory smoothing result is obtained by using the cubic spline curve smoothing result as [x smooth_i ,y smooth_i ,θ smooth_i ],i=1:n smooth Among them, n smooth Represents the discrete trajectory point data of the smoothed trajectory result, x smooth_i ,y smooth_i ,θ smooth_i Respectively represent the horizontal coordinate, vertical coordinate and heading angle of the ith point after smoothing. The preliminary result after cubic spline smoothing is used as the reference line of the driving path.
[0069] Based on the driving path reference line, the vehicle trajectory of the vehicle itself is optimized to obtain the vehicle trajectory optimization result. Specifically, a nonlinear optimization problem can be established to perform trajectory planning in complex scenarios.
[0070] The technical solution of the embodiment of the present invention is to construct an initial grid map according to the position information of the own vehicle and the obstacle information around the vehicle when it is detected that the own vehicle meets the judgment conditions for entering a complex scene; convert the resolution of the initial grid map according to the grid positions in the initial grid map and the vehicle grid positions of the own vehicle in the initial grid map to obtain a transformed grid map; binarize the transformed grid map to obtain a binary grid map; construct a vehicle collision detection model of the own vehicle, and perform path search in the binary grid map based on the vehicle collision detection model to obtain a driving path reference line; optimize the vehicle trajectory of the own vehicle based on the driving path reference line to obtain a vehicle trajectory optimization result. The above technical solution reduces the dependence on the reference path provided by the high-precision map and perception information, reduces the dependence on the manually labeled reference path information, enhances the traffic capacity of the pilot assisted driving function in complex scenes, and improves the reliability of vehicle trajectory optimization in complex scenes.
[0071] Embodiment 2
[0072] Figure 2A This is a flow chart of a vehicle trajectory optimization method provided in Embodiment 2 of the present invention. This embodiment is optimized and improved on the basis of the above technical solutions.
[0073] Further, the step of "binarizing the transformed raster map to obtain a binary raster map" is refined into "performing a distance transformation on the transformed raster map to obtain a distance raster map; the distance raster map includes at least two raster areas with different resolutions; and binarizing each raster area in the distance raster map to obtain a binary raster map." to improve the raster map binarization method.
[0074] The step of "optimizing the vehicle trajectory of the own vehicle based on the driving path reference line to obtain the vehicle trajectory optimization result" is refined into "constructing target constraints including vehicle kinematic constraints, vehicle starting state constraints and other vehicle operating state constraints; constructing a path optimization objective function according to the driving path reference line; and determining the vehicle trajectory optimization result of the own vehicle when the function value of the path optimization objective function is minimized when the target constraints are met." to improve the vehicle trajectory optimization method.
[0075] It should be noted that for the parts not described in detail in the embodiments of the present invention, reference can be made to the descriptions of other embodiments. Figure 2A As shown, the method comprises the following specific steps:
[0076] S210: When it is detected that the vehicle meets the conditions for entering a complex scene, an initial grid map is constructed according to the location information of the vehicle and the obstacle information around the vehicle.
[0077] S220 , performing resolution conversion on the initial grid map according to each grid position in the initial grid map and the vehicle grid position of the own vehicle in the initial grid map to obtain a transformed grid map.
[0078] S230, performing distance transformation on the transformed grid map to obtain a distance grid map; the distance grid map includes at least two grid areas with different resolutions.
[0079] In order to ensure the safety of planning results, the discrete grid map is transformed to calculate the distance between each grid in the transformed grid map and the nearest inaccessible area.
[0080] In an optional embodiment, a transformed grid map is subjected to a distance transformation to obtain a distance grid map, including: for any grid area in the transformed grid map, determining at least one occupied grid in the grid area; determining a reference grid distance between an unoccupied grid in the grid area and each occupied grid; selecting a reference grid distance that satisfies a shortest distance judgment condition as a target reference grid distance of the unoccupied grid; using the target reference grid distance in each grid area as a grid distance value of the unoccupied grid, and using a preset initial grid distance in each grid area as a grid distance value of the occupied grid to obtain a distance grid map.
[0081] For any grid area with different resolutions in the transformed grid map, the occupied grid in the grid area is determined, and the grid distance value of the occupied grid is updated to the initial grid distance value, that is, 0. Unoccupied grids other than the occupied grids are searched one by one in the grid area. For any unoccupied grid, the occupied grid with the shortest distance to the unoccupied grid is selected from the grid area, and the reference grid distance between the unoccupied grid and the occupied grid with the shortest distance is determined, and the reference grid distance is determined as the grid distance value of the unoccupied grid.
[0082] It should be noted that due to the different resolutions of each grid area, the grid distance values corresponding to the grids at the critical points between grid areas with different resolutions may not be accurate enough. Therefore, for the grids in the intersection grid area between grid areas with different resolutions, it is necessary to update the grid distance values of the grids that meet the conditions.
[0083] Optionally, determine the grids to be updated that meet the preset conditions for judging grids with similar distances between the grid regions; compare the grid distance values of the grids to be updated corresponding to the grid regions to determine whether to update the grid distance values of the grids to be updated.
[0084] For any two adjacent grid areas, a grid area with a close distance between the two grid areas is determined, and the grids in the grid area with a close distance are determined as the grids to be updated. The determination condition of the grids with a close distance may specifically be that the grid distance meets a certain threshold condition.
[0085] like Figure 2B A schematic diagram of a determination process of a grid to be updated is shown. Assuming that two grid regions with different resolutions are grid region A and grid region B, select any grid a to be updated from grid region A, and select a grid b to be updated that is closest to the grid a to be updated from grid region B, and compare the grid distance value of the grid a to be updated with the grid distance value of the grid b to be updated. If the grid distance value of the grid a to be updated is greater than the grid b to be updated, the grid distance value of the grid a to be updated is updated using the grid distance value of the grid b to be updated, and the grid distance value of the grid b to be updated remains unchanged. If the grid distance value of the grid a to be updated is less than the grid b to be updated, the grid distance value of the grid a to be updated is updated using the grid distance value of the grid b to be updated, and the grid distance value of the grid a to be updated remains unchanged. If the grid distance value of the grid a to be updated is equal to the updated grid b, the grid distance value of the grid b to be updated and the grid distance value of the grid a to be updated remain unchanged. Mark the grid a and the grid b to be updated as updated, continue to select the grid c to be updated that is not marked as updated from the grid area A, and select the grid b to be updated that is closest to the grid c to be updated from the grids to be updated that are not marked as updated in the grid area B, and repeat the process until all the grids to be updated are marked as updated, then end the cycle and the grid distance value update process.
[0086] S240 , performing binarization conversion on each grid area in the distance grid map to obtain a binary grid map.
[0087] Since different grid areas in the distance grid map have different resolutions, a binary conversion is performed based on different safety distance thresholds of grid areas with different resolutions to obtain a binary grid map.
[0088] In an optional embodiment, each grid area in the distance grid map is binarized to obtain a binary grid map, including: binarizing the distance grid map according to the safety distance thresholds corresponding to each grid area in the distance grid map to obtain a binary grid map.
[0089] The safety distance thresholds corresponding to the grid areas with different resolutions can be preset by relevant technicians based on actual experience or experimental values. The safety distance threshold is used to determine whether each grid is safe enough for passage.
[0090] Exemplarily, for any grid in any grid area, determine whether the grid distance value of the grid is greater than the safety distance threshold corresponding to the grid area; if so, update the grid binarization value of the grid area to 1, and mark the grid as passable; if not, update the grid binarization value of the grid area to 0, and mark the grid as inpassable.
[0091] S250: construct a vehicle collision detection model for the own vehicle, and perform a path search in a binary grid map based on the vehicle collision detection model to obtain a driving path reference line.
[0092] Establish a nonlinear optimization problem to plan the complex scene passage trajectory. Obtain the vehicle's position information, speed information, heading angle information, acceleration information, and steering wheel angle information. Select the following information as state variables in the planning problem:
[0093] X=[x i ,y i ,θ i ,k i ,v i ,a i ],i=1:n;
[0094] Among them, x i represents the lateral position coordinate of the vehicle at the i-th moment; y i represents the longitudinal position coordinate of the vehicle at the i-th moment; θ i represents the vehicle heading angle at the i-th moment, k i represents the curvature of the vehicle at the i-th moment, v i represents the vehicle speed at the i-th moment, a i represents the vehicle acceleration at the i-th moment.
[0095] The following information is selected as control variables:
[0096]
[0097] in, represents the rate of change of curvature over time at the i-th moment, j i It represents the rate of change of acceleration at the i-th moment.
[0098] S260: Construct target constraints including vehicle kinematic constraints, vehicle initial state constraints, and other vehicle operating state constraints.
[0099] Considering the vehicle kinematic constraints, the following vehicle kinematic constraints are constructed:
[0100]
[0101] Construct the vehicle's starting state constraints:
[0102]
[0103] Among them, x init Indicates the lateral coordinate position of the vehicle at the initial moment; y init represents the longitudinal coordinate position of the vehicle at the initial moment; θ init represents the vehicle heading angle at the initial moment; k init represents the curvature of the vehicle at the initial moment; v init represents the curvature of the vehicle at the initial moment; a init Represents the vehicle acceleration at the initial moment.
[0104] Taking into account the control quantity and vehicle speed and curvature limits, the following other vehicle operating state constraints are constructed:
[0105]
[0106] The upper and lower limits of each state variable and control variable are given in combination with the current vehicle operating environment and calibration results.
[0107] S270: Construct a path optimization objective function according to the driving path reference line.
[0108] Among them, the path optimization objective function is constructed as follows:
[0109]
[0110] Among them, w speed represents the speed deviation weight; w dis Indicates the deviation weight of the driving path reference line; represents the centripetal acceleration weight; w jerk is the acceleration change rate weight; w dkappa represents the rate of change of curvature; v target is the expected speed; d({x i ,y i},l ref ) is the vehicle position {x i ,y i}On the driving path reference line l ref projection distance.
[0111] S280. Under the condition that the target constraint condition is satisfied, determine the vehicle trajectory optimization result of the own vehicle when the function value of the path optimization target function is minimum.
[0112] The above problem is solved using a nonlinear optimization solver, and the cubic spline curve fitting result is used as the initial solution to improve the problem solving speed. Finally, the trajectory optimization result is extracted from the final result of the iterative optimization, and the optimized trajectory result is sent to the control module to be converted into a vehicle control command to enable the vehicle to pass through complex scenes. When the target constraint conditions are met and the function value of the path optimization objective function is minimized, the vehicle trajectory optimization result of the vehicle itself is as follows:
[0113] X res =[x i ,y i ,θ i ,k i ,v i ,a i ],i=1:n;
[0114] The technical solution of this embodiment performs distance transformation on the transformed grid map to obtain a distance grid map, and performs binarization transformation on each grid area in the distance grid map to obtain a binary grid map. In the process of distance transformation, the influence of resolution is fully considered, so that the distance grid map and the binary grid map obtained are more accurate, and the conversion and search efficiency are higher. In the process of trajectory optimization, a nonlinear optimization method is used to construct the optimization problem, and a synchronous search is performed in time and space. Compared with the trajectory search method separated in time and space, it has stronger environmental adaptability and dynamic environment interaction ability.
[0115] Embodiment 3
[0116] Figure 3 The schematic diagram of the structure of a vehicle trajectory optimization device provided in the third embodiment of the present invention. The vehicle trajectory optimization device provided in the embodiment of the present invention can be applied to the case of optimizing the trajectory of a vehicle entering a complex scene. The vehicle trajectory optimization device can be implemented in the form of hardware and / or software, such as Figure 3 As shown, the vehicle trajectory optimization device can be configured in an electronic device, and specifically includes: a map construction module 301, a resolution conversion module 302, a binary conversion module 303, a model construction module 304 and a trajectory optimization module 305. Among them,
[0117] A map construction module 301 is used to construct an initial grid map based on the vehicle position information and the obstacle information around the vehicle when the vehicle meets the judgment condition of entering a complex scene;
[0118] A resolution conversion module 302 is used to convert the resolution of the initial grid map according to the grid positions in the initial grid map and the vehicle grid position of the own vehicle in the initial grid map to obtain a transformed grid map;
[0119] A binary conversion module 303 is used to perform a binary conversion on the transformed grid map to obtain a binary grid map;
[0120] A model building module 304 is used to build a vehicle collision detection model of the own vehicle, and perform a path search in the binary grid map based on the vehicle collision detection model to obtain a driving path reference line;
[0121] The trajectory optimization module 305 is used to optimize the vehicle trajectory of the own vehicle based on the driving path reference line to obtain a vehicle trajectory optimization result.
[0122] The technical solution of the embodiment of the present invention is to construct an initial grid map according to the position information of the own vehicle and the obstacle information around the vehicle when it is detected that the own vehicle meets the judgment conditions for entering a complex scene; convert the resolution of the initial grid map according to the grid positions in the initial grid map and the vehicle grid positions of the own vehicle in the initial grid map to obtain a transformed grid map; binarize the transformed grid map to obtain a binary grid map; construct a vehicle collision detection model of the own vehicle, and perform path search in the binary grid map based on the vehicle collision detection model to obtain a driving path reference line; optimize the vehicle trajectory of the own vehicle based on the driving path reference line to obtain a vehicle trajectory optimization result. The above technical solution reduces the dependence on the reference path provided by the high-precision map and perception information, reduces the dependence on the manually labeled reference path information, enhances the traffic capacity of the pilot assisted driving function in complex scenes, and improves the reliability of vehicle trajectory optimization in complex scenes.
[0123] Optionally, the device further comprises:
[0124] The first scene judgment module is used to obtain complex scene distance information sent by the map navigation application, and if the complex scene distance information is less than a preset scene distance threshold, it is determined that the own vehicle meets the complex scene judgment condition; or,
[0125] The second scene judgment module is used to determine that the vehicle meets the judgment condition for entering a complex scene when a complex scene feature identification element is detected.
[0126] Optionally, the resolution conversion module 302 includes:
[0127] A first resolution adjustment unit is used to adjust the resolution corresponding to the first grid area in the initial grid map, where the grid position and the vehicle grid position meet the first judgment condition, to a first resolution; and
[0128] A second resolution adjustment unit is used to adjust the resolution corresponding to the second grid area in the initial grid map, where the grid position and the vehicle grid position meet the second judgment condition, to a second resolution; and
[0129] A third resolution adjustment unit, configured to adjust the resolution corresponding to a third grid area in the initial grid map, where the grid position and the vehicle grid position meet a third judgment condition, to a third resolution;
[0130] The transformed grid map generating unit is used to generate a transformed grid map including a first grid area of a first resolution, a second grid area of a second resolution, and a third grid area of a third resolution.
[0131] Optionally, the binarization conversion module 303 includes:
[0132] A distance transformation unit, used for performing distance transformation on the transformation grid map to obtain a distance grid map; the distance grid map includes at least two grid areas with different resolutions;
[0133] The binarization conversion unit is used to perform binarization conversion on each of the grid areas in the distance grid map to obtain a binarized grid map.
[0134] Optionally, the distance transformation unit includes:
[0135] An occupied grid determination subunit, for determining at least one occupied grid in any grid area in the transformed grid map;
[0136] A reference distance determination subunit, used to determine a reference grid distance between an unoccupied grid in the grid area and each of the occupied grids;
[0137] A target distance determination subunit, used to select a reference grid distance that meets the shortest distance judgment condition as the target reference grid distance of the unoccupied grid;
[0138] The distance grid map determination subunit is used to use the target reference grid distance in each grid area as the grid distance value of the unoccupied grid, and use the preset initial grid distance in each grid area as the grid distance value of the occupied grid to obtain a distance grid map.
[0139] Optionally, the distance transformation unit further includes:
[0140] The to-be-updated grid determination subunit is used to determine the to-be-updated grids that meet the preset close-distance grid determination condition between the grid regions;
[0141] The numerical comparison subunit is used to compare the grid distance values of the grids to be updated corresponding to the grid areas to determine whether to update the grid distance values of the grids to be updated.
[0142] Optional, binary conversion unit, including:
[0143] The binarization conversion subunit is used to perform binarization conversion on the distance grid map according to the safety distance thresholds respectively corresponding to the grid areas in the distance grid map to obtain a binarized grid map.
[0144] Optionally, the trajectory optimization module 305 includes:
[0145] A target constraint construction unit, used to construct target constraint conditions including vehicle kinematic constraints, vehicle initial state constraints and other vehicle operating state constraints;
[0146] An objective function construction unit, used to construct a path optimization objective function according to the driving path reference line;
[0147] The vehicle trajectory optimization unit is used to determine the vehicle trajectory optimization result of the own vehicle when the function value of the path optimization objective function is minimized while satisfying the target constraint condition.
[0148] The vehicle trajectory optimization device provided in the embodiment of the present invention can execute the vehicle trajectory optimization method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.
[0149] Embodiment 4
[0150] Figure 4 A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.
[0151] like Figure 4As shown, the electronic device 40 includes at least one processor 41, and a memory connected to the at least one processor 41, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 to the random access memory (RAM) 43. In the RAM 43, various programs and data required for the operation of the electronic device 40 can also be stored. The processor 41, the ROM 42, and the RAM 43 are connected to each other through a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.
[0152] A number of components in the electronic device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0153] The processor 41 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 41 executes the various methods and processes described above, such as the vehicle trajectory optimization method.
[0154] In some embodiments, the vehicle trajectory optimization method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 40 via the ROM 42 and / or the communication unit 49. When the computer program is loaded into the RAM 43 and executed by the processor 41, one or more steps of the vehicle trajectory optimization method described above may be performed. Alternatively, in other embodiments, the processor 41 may be configured to perform the vehicle trajectory optimization method in any other suitable manner (e.g., by means of firmware).
[0155] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0156] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0157] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0158] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).
[0159] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.
[0160] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.
[0161] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.
[0162] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A vehicle trajectory optimization method, characterized in that: include: When the vehicle meets the conditions for entering a complex scene, an initial grid map is constructed based on the vehicle's position information and the obstacle information around the vehicle. According to each grid position in the initial grid map and the vehicle grid position of the own vehicle in the initial grid map, the initial grid map is converted into a transformed grid map; Binarizing the transformed raster map to obtain a binary raster map; Constructing a vehicle collision detection model of the own vehicle, and performing a path search in the binary grid map based on the vehicle collision detection model to obtain a driving path reference line; Based on the driving path reference line, the vehicle trajectory of the own vehicle is optimized to obtain a vehicle trajectory optimization result.
2. The method according to claim 1, characterized in that The method further comprises: Obtaining complex scene distance information sent by a map navigation application, and if the complex scene distance information is less than a preset scene distance threshold, determining that the own vehicle meets the judgment condition for entering a complex scene; or, When a complex scene feature identification element is detected, it is determined that the vehicle meets the complex scene judgment condition.
3. The method according to claim 1, characterized in that The step of converting the resolution of the initial grid map according to the grid positions in the initial grid map and the vehicle grid position of the own vehicle in the initial grid map to obtain a transformed grid map comprises: adjusting the resolution corresponding to the first grid area in the initial grid map, where the grid position and the vehicle grid position meet the first judgment condition, to the first resolution; and adjusting the resolution corresponding to the second grid area in the initial grid map, where the grid position and the vehicle grid position meet the second judgment condition, to the second resolution; and Adjusting the resolution corresponding to the third grid area in the initial grid map, where the grid position and the vehicle grid position meet the third judgment condition, to a third resolution; A transformed grid map is generated that includes a first grid region at a first resolution, a second grid region at a second resolution, and a third grid region at a third resolution.
4. The method according to claim 1, characterized in that The step of performing a binarization conversion on the transformed grid map to obtain a binarized grid map includes: Performing distance transformation on the transformed grid map to obtain a distance grid map; the distance grid map includes at least two grid areas with different resolutions; Each of the grid areas in the distance grid map is binarized to obtain a binarized grid map.
5. The method according to claim 4, characterized in that The step of performing distance transformation on the transformed grid map to obtain a distance grid map comprises: For any grid area in the transformed grid map, determining at least one occupied grid in the grid area; Determine a reference grid distance between an unoccupied grid in the grid area and each of the occupied grids; Selecting a reference grid distance that satisfies the shortest distance judgment condition as the target reference grid distance of the unoccupied grid; The target reference grid distance in each of the grid areas is used as the grid distance value of the unoccupied grid, and the preset initial grid distance in each of the grid areas is used as the grid distance value of the occupied grid, so as to obtain a distance grid map.
6. The method according to claim 5, characterized in that The method further comprises: Determine the grids to be updated that meet the preset conditions for determining grids with similar distances between the grid areas; The grid distance values of the grids to be updated corresponding to the grid regions are numerically compared to determine whether to update the grid distance values of the grids to be updated.
7. The method according to claim 4, characterized in that The step of performing binarization conversion on each of the grid areas in the distance grid map to obtain a binarized grid map comprises: According to the safety distance thresholds respectively corresponding to the grid areas in the distance grid map, the distance grid map is binarized to obtain a binarized grid map.
8. The method according to claim 1, characterized in that The optimizing the vehicle trajectory of the own vehicle based on the driving path reference line to obtain the vehicle trajectory optimization result includes: Construct target constraints including vehicle kinematic constraints, vehicle initial state constraints and other vehicle operating state constraints; Constructing a path optimization objective function according to the driving path reference line; Under the condition that the target constraint condition is met, the vehicle trajectory optimization result of the own vehicle is determined when the function value of the path optimization objective function is minimized.
9. A vehicle trajectory optimization device, characterized in that: include: A map construction module is used to construct an initial grid map based on the vehicle's position information and the obstacle information around the vehicle when the vehicle meets the judgment conditions for entering a complex scene; A resolution conversion module, used for performing resolution conversion on the initial grid map according to each grid position in the initial grid map and a vehicle grid position of the own vehicle in the initial grid map to obtain a transformed grid map; A binary conversion module, used for performing a binary conversion on the transformed raster map to obtain a binary raster map; A model building module is used to build a vehicle collision detection model of the own vehicle, and perform a path search in the binary grid map based on the vehicle collision detection model to obtain a driving path reference line; The trajectory optimization module is used to optimize the vehicle trajectory of the own vehicle based on the driving path reference line to obtain a vehicle trajectory optimization result.
10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the vehicle trajectory optimization method according to any one of claims 1 to 8.
11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the vehicle trajectory optimization method according to any one of claims 1 to 8 when executed.
12. A computer program product, characterized in that The computer program product comprises a computer program, which, when executed by a processor, implements the vehicle trajectory optimization method according to any one of claims 1 to 8.
Citation Information
Cited By
Method and device for determining feasible region limitation scene, equipment, medium and program product
CN121115570A
