WO2022063005A1 - 全局路径规划方法及装置 - Google Patents

全局路径规划方法及装置 Download PDF

Info

Publication number
WO2022063005A1
WO2022063005A1 PCT/CN2021/118566 CN2021118566W WO2022063005A1 WO 2022063005 A1 WO2022063005 A1 WO 2022063005A1 CN 2021118566 W CN2021118566 W CN 2021118566W WO 2022063005 A1 WO2022063005 A1 WO 2022063005A1
Authority
WO
WIPO (PCT)
Prior art keywords
point
closest
edge
closest point
skeleton line
Prior art date
Application number
PCT/CN2021/118566
Other languages
English (en)
French (fr)
Inventor
张卓
颜波
徐成
张放
李晓飞
张德兆
王肖
霍舒豪
Original Assignee
北京智行者科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京智行者科技有限公司 filed Critical 北京智行者科技有限公司
Publication of WO2022063005A1 publication Critical patent/WO2022063005A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
    • G01C21/34—Route searching; Route guidance
    • G01C21/3446—Details of route searching algorithms, e.g. Dijkstra, A*, arc-flags, using precalculated routes
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Definitions

  • the present invention relates to the technical field of automatic driving, and in particular, to a global path planning method and device.
  • Unmanned driving technology can be briefly divided into perception, prediction, positioning, decision-making, planning and control.
  • Planning refers to path planning, which is mainly divided into global path planning and local path planning.
  • the goal of global path planning is to reasonably search the environmental area within the determined area, and after considering the safety threshold, generate an optimal collision-free path from the current position of the vehicle to any target position in the area.
  • unstructured regions lack information describing the topological relationship of paths, so it is more challenging to perform efficient global path planning in unstructured regions.
  • the global path planning scheme of unmanned driving is mainly divided into two types: grid-based graph search algorithm and random sampling-based search algorithm:
  • the grid-based graph search algorithm is represented by the A* algorithm and its variants.
  • the main idea is to rasterize the exploration area and explore the surrounding space with a grid as the sampling step. Each sampling point is evaluated and screened through the designed heuristic function, and the exploration continues until the target point is reached. After exploring to the target point, the final search path is obtained according to the "parent node" set during the exploration.
  • the search algorithm based on random sampling is represented by the RRT algorithm. Its main idea is to use the starting point as the root node, and generate a random expansion tree by randomly sampling and adding leaf nodes. When the leaf nodes in the random tree contain the target point or enter the target area, they can be found in the random tree. A path from the initial point to the target point.
  • the search algorithm based on random sampling does not need to rasterize the map in advance, it is a pure random search algorithm and is not sensitive to the type of environment.
  • the algorithm's performance Convergence is slow, efficiency drops significantly, and it is difficult to find paths in environments with narrow channels.
  • the purpose of the embodiments of the present invention is to provide a global path planning method and device, so as to solve the problem that the rationality of the grid-based search algorithm in the prior art to generate a path depends on the shape and accuracy of the grid, and needs to be based on the width and accuracy of the passing area.
  • the shape adopts different grid sizes and shapes, which will not only increase the cost of map production, but also increase the problem of computational complexity.
  • the search algorithm based on random sampling is not sensitive to the type of environment. When the search space contains a large number of obstacles or narrow passage constraints When , the convergence speed of the algorithm is slow, the efficiency will be greatly reduced, and it is difficult to find a path in an environment with narrow channels.
  • the present invention provides a global path planning method, and the global path planning method includes:
  • the unstructured road includes a passable area and an impassable area
  • a plurality of edges and a plurality of vertices of the unstructured road are determined, and the plurality of edges and the plurality of vertices form a class-structured map; the edges have Length attribute and direction attribute, the intersection of any two edges is the vertex of the unstructured road;
  • the first position of the vehicle's current position in the class-structured map and the second position of the target point in the class-structured map determine the first position on the edge of the class-structured map from the first position A set of closest points, and a second set of closest points on the edge of the second position to the class structured map;
  • each first closest point in the first closest point set to the sampling point along the edge determines The evaluation function value of each first closest point in the first closest point set to each second closest point in the second closest point set;
  • the obtaining the first skeleton line of the passable area in the unstructured road specifically includes:
  • the first skeleton line is determined according to the any grid value.
  • the obtaining the right driving path of the passable area specifically includes:
  • a right driving path is determined.
  • the grid size is determined according to the width of the vehicle
  • the determining of the second skeleton line according to the distance between each point in the first skeleton line and the boundary of the unstructured road specifically includes:
  • the point in the first skeleton line is deleted to obtain a second skeleton line.
  • the edge having a direction attribute specifically includes:
  • the first side point When the first side point is not within the unstructured road, determine the direction attribute of any side line as the first direction, and when the second side point is not within the unstructured road, determine the The direction property of any edge line is the second direction; the first direction and the second direction are opposite directions to each other.
  • the search is performed from the first position to the first closest point set on the edge of the structured map, and the second closest point from the second position to the edge of the structured map.
  • the point set specifically includes:
  • a second set of closest points in the end point circle is determined, and each second closest point in the second set of closest points is used as the end point of the solicitation path planning.
  • determining the evaluation value of each first closest point in the first closest point set to each second closest point in the second closest point set specifically includes:
  • an evaluation function value is determined.
  • the present invention provides a global path planning apparatus, the global path planning apparatus includes:
  • the obtaining unit is configured to obtain a first skeleton line of a passable area in an unstructured road;
  • the unstructured road includes a passable area and an impassable area;
  • a determining unit configured to determine a second skeleton line according to the distance between each point in the first skeleton line and the boundary of the unstructured road;
  • the obtaining unit is further configured to obtain the right-hand driving path of the passable area
  • the determining unit is further configured to, according to the second skeleton line and the right driving path, determine multiple edges and multiple vertices of the unstructured road, where the multiple edges and the multiple vertices form a class. Structured map; the edge has a length attribute and a direction attribute, and the intersection of any two edges is the vertex of the unstructured road;
  • mapping unit is used to map the current position of the vehicle and the target point to the class structured map before performing global path planning
  • the determining unit is further configured to determine the first position to the class structure according to the first position of the current position of the vehicle in the class structured map and the second position of the target point in the class structured map A first set of closest points on the edge of the structured map, and a second set of closest points on the edge of the second location to the class-structured map;
  • the determining unit is further configured to, according to the length of each first closest point in the first closest point set to the sampling point along the edge, the direction along the edge from each first closest point to the sampling point, and the sampling point to each The distance of the second closest point, determining the evaluation function value of each first closest point in the first closest point set to each second closest point in the second closest point set;
  • the determining unit is further configured to determine the path between the first closest point with the smallest evaluation function value to the second closest point as the optimal path.
  • the present invention provides a device including a memory and a processor, where the memory is used for storing a program, and the processor is used for executing any one of the methods in the first aspect.
  • the present invention provides a computer program product comprising instructions, when the computer program product is run on a computer, causing the computer to perform the method according to any one of the first aspects.
  • the present invention provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the method according to any one of the first aspects is implemented.
  • the present invention provides a computer server, comprising: a memory, a processor and a transceiver; the processor is configured to be coupled with the memory, read and execute instructions in the memory, so as to achieve the first The global path planning method according to any one of the aspects; the transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.
  • the present invention provides a mobile tool comprising the computer server of the sixth aspect.
  • the unstructured road can be topologically processed to obtain its quasi-structured map, so that the quasi-structured map can be used for global path planning to improve the efficiency of global path planning.
  • the present application can combine the length attribute and direction attribute of the edge in the structured-like map when evaluating the path, thereby improving the rationality of the path search result.
  • FIG. 1 is a schematic flowchart of a global path planning method according to Embodiment 1 of the present invention
  • Embodiment 1 of the present invention is a schematic diagram of an unstructured road provided by Embodiment 1 of the present invention.
  • Fig. 3 is the principle diagram of generating the class-structured map of unstructured road provided by Embodiment 1 of the present invention.
  • Fig. 4 is the block diagram of Fig. 3;
  • Embodiment 5 is a first skeleton line of a sub-structured road provided by Embodiment 1 of the present invention.
  • FIG. 6 is a schematic diagram of a second skeleton line and a right-hand driving path provided by Embodiment 1 of the present invention.
  • Embodiment 7 is a schematic diagram of a direction attribute of an edge provided by Embodiment 1 of the present invention.
  • FIG. 8 is a schematic diagram of an unstructured road with a directional attribute according to Embodiment 1 of the present invention.
  • FIG. 9 is a schematic diagram of the first closest point set and the second closest point set provided in Embodiment 1 of the present invention.
  • FIG. 10 is a schematic structural diagram of a global path planning apparatus according to Embodiment 2 of the present invention.
  • FIG. 1 is a schematic flowchart of a global path planning method according to Embodiment 1 of the present invention.
  • the execution body of this application is a terminal, server or processor with computing functions.
  • This application takes the application of the method to an unmanned vehicle as an example for description.
  • the execution subject of the method is the Automatic Vehicle Control Unit (AVCU), that is, The central processing unit of a driverless vehicle is equivalent to the “brain” of the driverless vehicle.
  • AVCU Automatic Vehicle Control Unit
  • This application includes the following steps:
  • Step 110 Obtain the first skeleton line of the passable area in the unstructured road.
  • structured roads when the autonomous vehicle is driving, it will load a map of the driving area, and in the map, there are structured roads and unstructured roads.
  • unstructured roads are opposed to structured roads.
  • Structured roads refer to highways, urban arterial roads and other well-structured roads. Such roads have clear road marking lines, and the background environment of the road is relatively simple. The geometric features are also more obvious. For structured roads, it is easy to obtain the topological relationship between paths, and then perform global path planning according to the topological relationship.
  • Unstructured roads generally refer to roads with a low degree of structure such as urban non-arterial roads and rural streets. Such roads do not have lane lines and clear road boundaries. In addition, they are affected by shadows and water marks. Non-road areas are indistinguishable. Unstructured roads include passable areas and impassable areas.
  • FIG. 2 is a schematic diagram of an unstructured road provided by Embodiment 1 of the present invention.
  • the lower point in FIG. 2 can be used as the current position of the vehicle, and the upper point can be used as the target point of the current vehicle. Dark areas are passable areas and white areas are impassable areas.
  • a certain process can be used to determine its first skeleton line.
  • the processing here is rasterization processing.
  • the steps of determining the first skeleton line through rasterization processing are as follows:
  • the passable area in the unstructured road is rasterized, and the passable area is divided into multiple grids; secondly, according to the distance from each grid to the impassable area, the grid of each grid is determined grid value; finally, when any grid value represents a different part in the distance impassable area, the first skeleton line is determined according to any grid value.
  • the following describes how to determine the first skeleton line with reference to FIG. 3 .
  • the black area in Figure 3 is the impassable area
  • the gray area is the passable area
  • the impassable area is regarded as an obstacle
  • the boundary of the unstructured road in Figure 3 is regarded as a separate obstacle
  • the passable area is fenced Grid processing, each grid value represents the distance to the nearest obstacle, the grid value of the grid occupied by the obstacle is 0, the grid value of the grid adjacent to the obstacle is recorded as 1, and it extends outward in turn.
  • the grid value of 4 means that the grid value is 4 by two Different obstacles extend to the same grid, and at this time, the grid at the end of the extension constitutes the first skeleton line of the area.
  • Fig. 4 is the block diagram of Fig. 3.
  • the passable area can be divided into 5 blocks according to the white line.
  • the nearest obstacle to the point of area 1 is the obstacle on the left
  • the obstacles on the other sides are not as close as the obstacles on the left.
  • the closest obstacle to Zone 2 is the obstacle on the upper side
  • the obstacles on the other sides are not as close as the obstacles on the upper side.
  • the closest obstacle to Zone 5 is the middle obstacle, and the obstacles on the other sides are not as close as the middle obstacle.
  • connection between the grid and the vertex of the grid value 4 is used to distinguish different blocks.
  • the reason why the grid with grid value 4 takes the center point is similar, because the distance from the center point to the left, upper side, and middle obstacles is the same, so it is used as the dividing point.
  • the first skeleton line of the unstructured road shown in Fig. 2 can be determined.
  • the first skeleton line of the unstructured road shown in Fig. 2 is shown in Fig. 5, and Fig. 5 is The first skeleton line of the sub-structured road provided by the first embodiment of the present invention.
  • the grid size in the above-mentioned grid processing is determined according to the width of the vehicle.
  • the size of the grid and the width of the vehicle are in a fixed ratio, such as 1:1, or the size of the grid is Determined according to the length and width of the vehicle, for example, the size of the grid is the square of the length of the vehicle plus the square of the width, and then squared, so as to determine the size of the grid according to the length and width of the vehicle, which is convenient for improving grid processing. speed.
  • Step 120 Determine the second skeleton line according to the distance between each point in the first skeleton line and the boundary of the unstructured road.
  • the first skeleton line which is composed of multiple edge lines
  • the first skeleton line Delete the edge in the skeleton to get the second skeleton line.
  • the edge is eliminated.
  • Step 130 Obtain the right-hand driving path of the passable area.
  • the distance from the path when driving on the right to the boundary of the unstructured road can be determined according to the width of the unmanned vehicle and the minimum turning radius, so that the distance and the boundary of the unstructured road can be determined according to the distance and the boundary of the unstructured road. , determine the right driving path.
  • the distance here can be the sum of half the width of the vehicle and the minimum turning radius.
  • the grid adjacent to the impassable area may be obtained; the right driving path may be determined based on the grid values of the grid adjacent to the impassable area. For example, referring to Fig. 3, the grid value adjacent to the impassable area is 1, then according to the grid value 1, the center point of the grid with the grid value 1 is taken to connect a line, so as to determine the right driving path.
  • Step 140 Determine multiple edges and multiple vertices of the unstructured road according to the second skeleton line and the right driving path, and multiple edges and multiple vertices form a structured map; the edge has a length attribute and a direction attribute, and any The intersection of the two edges is the vertex of the unstructured road.
  • a structured-like map is actually similar to a structured map.
  • unstructured roads are treated as containing multiple edges and multiple vertices.
  • FIG. 6 is a schematic diagram of the second skeleton line and the right driving path provided by the first embodiment of the present invention
  • the solid line in FIG. 6 is the second skeleton line
  • the dotted line is the right driving path
  • the second skeleton in FIG. 6 The line and the right driving path are called edges.
  • the intersection of each edge is called a vertex. Between two vertices is an edge.
  • the edge has not only the length attribute, but also the direction attribute.
  • the second skeleton line in the edge is bidirectional. Passable, the right-hand driving path follows the principle of driving on the right.
  • any side point on any sideline on the right driving path first obtain any point on any sideline on the right driving path; then determine whether the first side point and the second side point of any point are in the unstructured road; the first side point and the second side point It is set on both sides of any point; finally, when the first side point is not in the unstructured road, the direction attribute of any side line is determined as the first direction, and when the second side point is not in the unstructured road, any side point is determined.
  • the direction property of a side line is the second direction; the first direction and the second direction are opposite directions.
  • CC r is the length of the line segment from point C to point Cr
  • CC l is the length of the line segment from point C to point C l
  • ⁇ is a value small enough to be considered infinitesimal.
  • the directional properties of edge AB can be determined by detecting whether C l and C r are within an unstructured road. If C r is not in the unstructured road, the arrow in Figure 8 is the direction of the AB edge; if C l is not in the unstructured road, the opposite direction of the arrow in Figure 8 is the direction of the AB edge.
  • the quasi-structured map of the unstructured road includes the length attribute, direction attribute and multiple vertices of the edge.
  • the direction attribute can be visually represented by an arrow, and the length attribute can be stored in the database.
  • an edge has an edge ID, and the corresponding edge ID corresponds to the length attribute of the edge.
  • the edge ID and length attributes can be stored in the database in a tabular fashion.
  • the ID of the sideline AB is 01 and the corresponding length is 5m, then 01 and 5m are stored in the database in the form of a table.
  • Step 150 before performing the global path planning, map the current position of the vehicle and the target point to the class structured map.
  • the current position information of the vehicle can be obtained through sensors installed on the vehicle, such as an inertial measurement unit (IMU), a global positioning system (Global Positioning System, GPS), etc.
  • IMU inertial measurement unit
  • GPS Global Positioning System
  • the server When the server receives the car-hailing service information from the user terminal, for example, the user terminal receives the information such as the departure place, destination and departure time entered by the user, and after receiving the departure place and destination information sent by the user terminal, the server filters out the target vehicle , and send the information such as departure place, destination, departure time, and number of departures to the target vehicle together.
  • the target vehicle when the vehicle is driving, it is necessary to carry out global path planning, so as to drive from the starting point to the destination.
  • the current position of the vehicle and the target point can be mapped in the class-structured map.
  • the current position of the vehicle can be latitude and longitude data
  • each point in the structured map has location information
  • the current latitude and longitude data of the vehicle can be mapped to the structured map, so that the vehicle's position in the structured map can be determined.
  • Current position and target point can be determined.
  • Step 160 according to the first position of the current position of the vehicle in the class-structured map and the second position of the target point in the class-structured map, determine the first closest point set on the edge line from the first position to the class-structured map , and the second closest point set on the edge of the second position to the class structured map.
  • step 160 includes: first, taking the first position as the center of the circle, and determining the starting circle with the preset first radius as the radius; secondly, determining the first closest point set in the starting circle, and collecting the first closest points Each first closest point of , is used as the starting point of the global path planning; again, take the second position as the center of the circle and the preset second radius as the radius to determine the end point circle; finally, determine the second closest point set in the end point circle, and set the Each second closest point in the second closest point set serves as the end point of the solicitation path planning.
  • the preset first radius and the preset second radius may be the same or different.
  • the preset first radius may be determined according to the width of the vehicle and a fixed value, and the fixed value is related to the number of sidelines around the first position and the distance between adjacent sidelines.
  • the first position is the lower Small rectangular frame, there are three edges at the first position, after making a circle with the preset first radius, make sure that the three edges are all in the starting circle, and the obtained first closest point set is below the three edges around the small rectangular frame.
  • a small five-pointed star indicates.
  • the setting of the end circle is similar to the setting of the start circle, which will not be repeated here.
  • the second position is the small rectangular frame above, and there are three borders at the first position.
  • the obtained second closest point set is represented by three small five-pointed stars around the small rectangular frame above. .
  • Step 170 according to the length of each first closest point in the first closest point set to the sampling point along the edge, the direction of each first closest point to the sampling point along the edge, and the distance from the sampling point to each second closest point, An evaluation function value for each of the first closest point in the first closest point set to each second closest point in the second closest point set is determined.
  • step 170 includes: determining the length along the edge from each starting point to the sampling point; determining the direction along the edge from each starting point to the sampling point; determining the first product of the preset first weight value and the length; The second product of the second weight value and the direction; determine the sum of the first product and the second product as the first generation value; determine the distance from the sampling point to the target point as the second generation value; according to the first generation value and the second generation value Cost value, determine the evaluation function value.
  • the sampling point is a point when each path search is performed, and the path search can be implemented by using a path search algorithm.
  • the A* algorithm in the heuristic path search algorithm, etc. is not limited in this application.
  • the evaluation can be performed through the evaluation function corresponding to the A* algorithm, and the evaluation function is shown in formula (2):
  • the value of g is the first generation value of the search, which represents the cost of searching from the starting point of the search to the current sampling point.
  • the h value is the second-generation value from the current sampling point to the target point, which can be represented by the distance from the sampling point to the target point.
  • the distance here includes but is not limited to the Euclidean distance and the Manhattan distance.
  • the present application can calculate the first generation value by formula (3).
  • l represents the length along the edge from the starting point to the current sampling point.
  • the edge is divided by vertices. The edge between two vertices is used as an edge, and dir is the direction from the starting point to the sampling point along the edge.
  • the forward line is recorded as 0
  • the retrograde is recorded as 1
  • ⁇ is the first weight coefficient, indicating the importance of the length of the path to the entire road
  • ⁇ is the second weight coefficient, indicating the importance of the direction to the entire path.
  • ⁇ and ⁇ can be set according to requirements. For example, if the vehicle is not required to travel in the opposite direction, but only to reach the target point in the shortest time, then ⁇ is of little significance. At this time, the value of ⁇ can be set to be smaller than the value of ⁇ . , and ⁇ is infinitely small, and its specific value needs to be determined through multiple simulation calculations.
  • Step 180 Determine the path between the first closest point with the smallest evaluation function value to the second closest point as the optimal path.
  • the evaluation function value may be determined according to the first generation value and the second generation value, and the corresponding optimal path with the smallest evaluation function value may be determined according to the evaluation function value.
  • the first generation of the three paths 1-A, 2-A, and 3-A is calculated from the starting point to the sampling point.
  • Value, from the sampling point to the end point calculate the second generation value of the three paths of A-4, A-5 and A-6, and then add them arbitrarily, that is, 1-A+A-4, 1-A+A-5, 1 -A+A-6, 2-A+A-4, 2-A+A-5, 2-A+A-6, 3-A+A-4, 3-A+A-5, 3-A +A-6, calculate these 9 values, and determine that the value of the evaluation function is the smallest as a local path.
  • 1-A+A-4 is the smallest, and the optimal path can be determined from the starting point 1 to the sampling point A and then to the end point 4.
  • the unstructured road can be topologically processed to obtain its quasi-structured map, so that the global path planning is carried out through the quasi-structured map, so as to improve the efficiency of the global path planning,
  • the present application can combine the length attribute and direction attribute of the edge in the structured-like map when performing the path evaluation, thereby improving the rationality of the path search result.
  • FIG. 10 is a schematic structural diagram of a global path planning apparatus according to Embodiment 2 of the present invention. As shown in FIG. 10 , the global path planning apparatus is applied to the global path planning method in Embodiment 1.
  • the global path planning apparatus includes: an obtaining unit 210 , a determining unit 220 and a mapping unit 230 .
  • the obtaining unit 210 is configured to obtain the first skeleton line of the passable area in the unstructured road; the unstructured road includes a passable area and an impassable area;
  • the determining unit 220 is configured to determine the second skeleton line according to the distance between each point in the first skeleton line and the boundary of the unstructured road;
  • the obtaining unit 210 is further configured to obtain the right driving path of the passable area
  • the determining unit 220 is further configured to, according to the second skeleton line and the driving path on the right, determine multiple edges and multiple vertices of the unstructured road, and multiple edge lines and multiple vertices form a class-structured map; the edge has a length attribute and Direction attribute, the intersection of any two edges is the vertex of the unstructured road;
  • the mapping unit 230 is used to map the current position of the vehicle and the target point to the class structured map before performing global path planning;
  • the determining unit 220 is further configured to, according to the first position of the vehicle's current position in the class-structured map and the second position of the target point in the class-structured map, determine the first position on the edge of the class-structured map. a set of closest points, and a second set of closest points on the edge of the second position to the class structured map;
  • the determining unit 220 is further configured to, according to the length of each first closest point in the first closest point set to the sampling point along the sideline, the direction along the sideline from each first closest point to the sampling point, and the distance from the sampling point to each second closest point. The distance of the points, determine the evaluation function value of each first closest point in the first closest point set to each second closest point in the second closest point set;
  • the determining unit 220 is further configured to determine the path between the first closest point with the smallest evaluation function value and the second closest point as the optimal path.
  • the acquiring unit 210 is specifically configured to: perform grid processing on the passable area in the unstructured road, and divide the passable area into multiple grids; determine the distance from each grid to the impassable area. The grid value of each grid; when any grid value represents a different part in the distance impassable area, the first skeleton line is determined according to any grid value.
  • the acquiring unit 210 is specifically configured to: acquire grids adjacent to the impassable area; and determine a right-hand driving path according to grid values of the grids adjacent to the impassable area.
  • the grid size is determined according to the width of the vehicle; the distance from the right driving path to the boundary of the unstructured road is determined according to the width of the vehicle and the minimum turning radius.
  • determining unit 220 is specifically used for:
  • the point in the first skeleton line is deleted to obtain the second skeleton line.
  • the edge has a direction attribute specifically including:
  • first side point When the first side point is not in the unstructured road, determine the direction attribute of any side line as the first direction, and when the second side point is not in the unstructured road, determine the direction attribute of any side line as the second direction;
  • the first direction and the second direction are opposite directions to each other.
  • determining unit 220 is specifically used for:
  • a second set of closest points in the end point circle is determined, and each second closest point in the second set of closest points is used as the end point of the solicitation path planning.
  • determining unit 220 is specifically used for:
  • Second product determine the sum of the first product and the second product as the first generation value; determine the distance from the sampling point to the target point as the second generation value; determine the evaluation function value according to the first generation value and the second generation value .
  • the unstructured road can be topologically processed to obtain its quasi-structured map, so that the global path planning is carried out through the quasi-structured map, so as to improve the efficiency of the global path planning,
  • the present application can combine the length attribute and direction attribute of the edge in the structured-like map when performing the path evaluation, thereby improving the rationality of the path search result.
  • Embodiment 3 of the present invention provides a device including a memory and a processor, where the memory is used to store a program, and the memory can be connected to the processor through a bus.
  • the memory may be non-volatile memory, such as hard drives and flash memory, in which software programs and device drivers are stored.
  • the software program can perform various functions of the above methods provided by the embodiments of the present invention; the device driver may be a network and interface driver.
  • the processor is configured to execute a software program, and when the software program is executed, the method provided by Embodiment 1 of the present invention can be implemented.
  • the fourth embodiment of the present invention provides a computer program product containing instructions, when the computer program product runs on a computer, the computer enables the computer to execute the method provided by the first embodiment of the present invention.
  • Embodiment 5 of the present invention provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the method provided by Embodiment 1 of the present invention is implemented.
  • Embodiment 6 of the present invention provides a computer server, including: a memory, a processor, and a transceiver; the processor is configured to be coupled with the memory, and read and execute instructions in the memory, so as to implement the first embodiment The global path planning method; the transceiver is coupled with the processor, and the processor controls the transceiver to send and receive messages.
  • the seventh embodiment of the present invention provides a mobile tool, including the computer server according to the sixth embodiment.
  • the above mobile tools can be any movable tools, such as vehicles (such as vacuum cleaners, sweepers, floor washing vehicles, logistics trolleys, passenger cars, sanitation vehicles, buses, coaches, vans, trucks, trucks, etc.) , trailer, dump trailer, crane, excavator, scraper, road train, sweeper, sprinkler, garbage truck, engineering vehicle, rescue vehicle, AGV (Automated Guided Vehicle), motorcycle, bicycle , tricycles, hand carts, robots, sweepers, balance vehicles, etc., this application does not strictly limit the types of mobile tools, and will not be exhaustive here.
  • vehicles such as vacuum cleaners, sweepers, floor washing vehicles, logistics trolleys, passenger cars, sanitation vehicles, buses, coaches, vans, trucks, trucks, etc.
  • a software module can be placed in random access memory (RAM), internal memory, read only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other in the technical field. in any other known form of storage medium.
  • RAM random access memory
  • ROM read only memory
  • electrically programmable ROM electrically erasable programmable ROM
  • registers hard disk, removable disk, CD-ROM, or any other in the technical field. in any other known form of storage medium.

Abstract

一种全局路径规划方法，包括：获取非结构化道路中的可通行区域的第一骨架线(S110)并据此确定第二骨架线(S120)；获取可通行区域的靠右行驶路径(S130)；根据第二骨架线和靠右行驶路径，确定非结构化道路的多条边线和多个顶点(S140)；边线具有长度属性和方向属性；在进行全局路径规划前，将车辆当前位置和目标点映射到类结构化地图上(S150)；确定第一位置到类结构化地图的边线上的第一最近点集，以及第二位置到类结构化地图的边线上的第二最近点集(S160)，确定第一最近点集中的每个第一最近点至第二最近点集中的每个第二最近点的评价函数值(S170)；确定评价函数值最小的第一最近点至第二最近点之间的路径为最优路径(S180)。方法提高了全局路径规划的效率且提升了路径搜索结果的合理性。还提供了一种全局路径规划装置、一种计算机可读存储介质、一种计算机服务器以及一种移动工具。

Description

全局路径规划方法及装置
本申请要求于2020年9月22日提交的、申请号为202011003654.1、标题为“全局路径规划方法及装置”的中国专利申请的优先权，该中国专利申请的公开内容以引用的方式并入本文。
技术领域
本发明涉及自动驾驶技术领域，尤其涉及一种全局路径规划方法及装置。
背景技术
随着人工智能技术以及现代制造业的发展，自动驾驶技术已经逐渐走进人们的日常生活，并潜移默化的改变着人们的出行方式。无人驾驶技术可以简要的分为感知、预测、定位、决策、规划与控制几个方面。规划指的是路径规划，主要分为全局路径规划和局部路径规划。全局路径规划的目标是在确定区域内，合理搜索环境区域，考虑安全阈值后，生成从车辆当前位置到区域内任意目标位置的一条最优无碰撞路径。相比于结构化区域，非结构化区域缺少描述路径拓扑关系的信息，所以在非结构化区域内进行高效的全局路径规划更加具有挑战性。
目前无人驾驶全局路径规划方案主要分为基于栅格的图搜索算法以及基于随机采样的搜索算法两种：
1.基于栅格的图搜索算法以A*算法以及其变种算法为代表。其主要思想是将探索区域进行栅格化，以一个栅格为采样步长向周围空间探索。通过设计的启发式函数对每次采样点进行评价与筛选，择优后继续探索直至探索到目标点。在探索至目标点后，依据探索时设定的“父节点”得到最终的搜索路径。
2.基于随机采样的搜索算法以RRT算法为代表。其主要思想它以起始点 作为根节点，通过随机采样增加叶子节点的方式，生成一个随机扩展树，当随机树中的叶子节点包含了目标点或进入了目标区域，便可以在随机树中找到一条由从初始点到目标点的路径。
基于栅格的搜索算法生成路径的合理性依赖栅格的形状和精度。对于范围大、路况复杂的非结构化区域，需要根据通行区域的宽度和形状采用不同的栅格大小和形状，不仅会增加地图的制作成本，也会增加计算量。
基于随机采样的搜索算法虽然不需要预先将地图进行栅格化，但是这是一种纯粹的随机搜索算法，对环境类型不敏感，当搜索空间中包含大量障碍物或狭窄通道约束时，算法的收敛速度慢，效率会大幅下降，而且难以在有狭窄通道的环境找到路径。
发明内容
本发明实施例的目的是提供一种全局路径规划方法及装置，以解决现有技术中的基于栅格的搜索算法生成路径的合理性依赖栅格的形状和精度，需要根据通行区域的宽度和形状采用不同的栅格大小和形状，不仅会增加地图的制作成本，也会增加计算量的问题以及基于随机采样的搜索算法对环境类型不敏感，当搜索空间中包含大量障碍物或狭窄通道约束时，算法的收敛速度慢，效率会大幅下降，而且难以在有狭窄通道的环境找到路径的问题。
第一方面，本发明提供了一种全局路径规划方法，所述全局路径规划方法包括：
获取非结构化道路中的可通行区域的第一骨架线；所述非结构化道路包括可通行区域和不可通行区域；
根据所述第一骨架线中的每个点与所述非结构化道路的边界的距离，确定第二骨架线；
获取所述可通行区域的靠右行驶路径；
根据所述第二骨架线和所述靠右行驶路径，确定非结构化道路的多条边 线和多个顶点，所述多条边线和所述多个顶点组成类结构化地图；所述边线具有长度属性和方向属性，任意两条边线的交点为非结构化道路的顶点；
在进行全局路径规划前，将车辆当前位置和目标点映射到所述类结构化地图上；
根据车辆当前位置在所述类结构化地图中的第一位置，以及目标点在所述类结构化地图中的第二位置，确定所述第一位置到类结构化地图的边线上的第一最近点集，以及所述第二位置到类结构化地图的边线上的第二最近点集；
根据所述第一最近点集中的每个第一最近点到采样点沿边线的长度、每个第一最近点到采样点沿边线的方向以及采样点至每个第二最近点的距离，确定第一最近点集中的每个第一最近点至第二最近点集中的每个第二最近点的评价函数值；
确定评价函数值最小的第一最近点至第二最近点之间的路径为最优路径。
在一种可能的实现方式中，所述获取非结构化道路中的可通行区域的第一骨架线具体包括:
对所述非结构化道路中的可通行区域进行栅格化处理，将所述可通行区域划分为多个栅格；
根据每个栅格至所述不可通行区域的距离，确定每个栅格的栅格值；
当任一栅格值表示的为距离所述不可通行区域中的不同部分时，根据所述任一栅格值，确定第一骨架线。
在一种可能的实现方式中，所述获取所述可通行区域的靠右行驶路径具体包括：
获取与所述不可通行区域相邻的栅格；
根据与所述不可通行区域相邻的栅格的栅格值，确定靠右行驶路径。
在一种可能的实现方式中，进行栅格化处理时，根据车辆的宽度确定栅格大小；
根据车辆的宽度和最小转弯半径，确定靠右行驶路径到非结构化道路的边界的距离。
在一种可能的实现方式中，所述根据所述第一骨架线中的每个点与所述非结构化道路的边界的距离，确定第二骨架线具体包括：
当所述第一骨架线中的点距离所述可通行区域的边界的距离小于预设的距离阈值时，将所述第一骨架线中的点删除，得到第二骨架线。
在一种可能的实现方式中，所述边线具有方向属性具体包括：
获取靠右行驶路径上的任一边线上的任意一点；
判断所述任意一点的第一侧点和第二侧点是否在非结构化道路内；所述第一侧点和所述第二侧点设置在所述任意一点的两侧；
当所述第一侧点不在所述非结构化道路内时，确定所述任一边线的方向属性为第一方向，当所述第二侧点不在所述非结构化道路内时，确定所述任一边线的方向属性为第二方向；所述第一方向和所述第二方向互为反方向。
在一种可能的实现方式中，所述搜索所述第一位置到类结构化地图的边线上的第一最近点集，以及所述第二位置到类结构化地图的边线上的第二最近点集具体包括：
以所述第一位置为圆心，以预设第一半径为半径确定起始圆；
确定所述起始圆中的第一最近点集，并将所述第一最近点集中的每一第一最近点作为全局路径规划的起点；
以所述第二位置为圆心，以预设第二半径为半径确定终点圆；
确定所述终点圆中的第二最近点集，并将所述第二最近点集中的每一第二最近点作为劝募路径规划的终点。
在一种可能的实现方式中，所述根据所述第一最近点集中的每个第一最近点到采样点沿边线的长度、每个第一最近点到采样点沿边线的方向以及采样点至每个第二最近点的距离，确定第一最近点集中的每个第一最近点至第二最近点集中的每个第二最近点的评价值具体包括：
确定每个起点到采样点沿边线的长度；
确定每个起点到采样点沿边线的方向；
确定预设的第一权重值和所述长度的第一乘积；
确定预设的第二权重值和所述方向的第二乘积；
确定所述第一乘积和所述第二乘积的和值作为第一代价值；
确定所述采样点到目标点的距离作为第二代价值；
根据所述第一代价值和所述第二代价值，确定评价函数值。
第二方面，本发明提供了一种全局路径规划装置，所述全局路径规划装置包括：
获取单元，所述获取单元用于获取非结构化道路中的可通行区域的第一骨架线；所述非结构化道路包括可通行区域和不可通行区域；
确定单元，所述确定单元用于根据所述第一骨架线中的每个点与所述非结构化道路的边界的距离，确定第二骨架线；
所述获取单元还用于，获取所述可通行区域的靠右行驶路径；
所述确定单元还用于，根据所述第二骨架线和所述靠右行驶路径，确定非结构化道路的多条边线和多个顶点，所述多条边线和所述多个顶点组成类结构化地图；所述边线具有长度属性和方向属性，任意两条边线的交点为非结构化道路的顶点；
映射单元，所述映射单元用于在进行全局路径规划前，将车辆当前位置和目标点映射到所述类结构化地图上；
所述确定单元还用于，根据车辆当前位置在所述类结构化地图中的第一位置，以及目标点在所述类结构化地图中的第二位置，确定所述第一位置到类结构化地图的边线上的第一最近点集，以及所述第二位置到类结构化地图的边线上的第二最近点集；
所述确定单元还用于，根据所述第一最近点集中的每个第一最近点到采样点沿边线的长度、每个第一最近点到采样点沿边线的方向以及采样点至每 个第二最近点的距离，确定第一最近点集中的每个第一最近点至第二最近点集中的每个第二最近点的评价函数值；
所述确定单元还用于，确定评价函数值最小的第一最近点至第二最近点之间的路径为最优路径。
第三方面，本发明提供了一种设备，包括存储器和处理器，存储器用于存储程序，处理器用于执行第一方面任一所述的方法。
第四方面，本发明提供了一种包含指令的计算机程序产品，当计算机程序产品在计算机上运行时，使得计算机执行如第一方面任一所述的方法。
第五方面，本发明提供了一种计算机可读存储介质，计算机可读存储介质上存储有计算机程序，计算机程序被处理器执行时实现如第一方面任一所述的方法。
第六方面，本发明提供了一种计算机服务器，包括：存储器、处理器和收发器；所述处理器用于与所述存储器耦合，读取并执行所述存储器中的指令，以实现如第一方面任一所述的全局路径规划方法；所述收发器与所述处理器耦合，由所述处理器控制所述收发器进行消息收发。
第七方面，本发明提供一种移动工具，包括如第六方面的计算机服务器。
通过应用本发明实施例提供的全局路径规划方法及装置，可以将非结构化道路进行拓扑化处理，得到其类结构化地图，从而通过类结构化地图经全局路径规划，以提高全局路径规划的效率，同时，本申请在进行路径评价时，可以结合类结构化地图中的边线的长度属性和方向属性，从而提升了路径搜索结果的合理性。
附图说明
图1为本发明实施例一提供的全局路径规划方法流程示意图；
图2为本发明实施例一提供的非结构化道路示意图；
图3为本发明实施例一提供的生成非结构化道路的类结构化地图的原理 图；
图4为图3的区块图；
图5为本发明实施例一提供的分结构化道路的第一骨架线；
图6为本发明实施例一提供的第二骨架线及靠右行驶路径示意图；
图7为本发明实施例一提供的边线的方向属性示意图；
图8为本发明实施例一提供的具有方向属性的非结构化道路的示意图；
图9为本发明实施例一提供的第一最近点集和第二最近点集示意图；
图10为本发明实施例二提供的全局路径规划装置结构示意图。
具体实施方式
下面结合附图和实施例对本申请作进一步的详细说明。可以理解的是，此处所描述的具体实施例仅仅用于解释相关发明，而非对该发明的限定。另外还需要说明的是，为便于描述，附图中仅示出了与有关发明相关的部分。
需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。下面将参考附图并结合实施例来详细说明本申请。
图1为本发明实施例一提供的全局路径规划方法流程示意图。本申请的执行主体为具有计算功能的终端、服务器或者处理器。本申请以将该方法应用在无人驾驶车辆为例进行说明，当将该方法应用在无人驾驶车辆时，该方法的执行主体为自动驾驶车辆控制单元(Automated Vehicle Control Unit，AVCU)，即无人驾驶车辆的中央处理器相当于无人驾驶车辆的“大脑”。本申请包括以下步骤：
步骤110，获取非结构化道路中的可通行区域的第一骨架线。
具体的，自动驾驶车辆在行驶中，会加载所行驶的区域的地图，该地图中，存在着结构化道路和非结构化道路。其中，非结构化道路与结构化道路相对立，结构化道路指的是高速公路、城市干道等结构化较好的公路，这类道路具有清晰的道路标志线，道路的背景环境比较单一，道路的几何特征也 比较明显。对于结构化道路，可以很容易的获取路径之间的拓扑关系，然后根据拓扑关系进行全局路径规划。
非结构化道路一般是指城市非主干道、乡村街道等结构化程度较低的道路，这类道路没有车道线和清晰的道路边界，再加上受阴影和水迹等的影响，道路区域和非道路区域难以区分。非结构化道路包括可通行区域和不可通行区域。
参见图2，图2为本发明实施例一提供的非结构化道路示意图，图2中下部的点可以作为车辆的当前位置，上部的点作为当前车辆的目标点。深色区域为可通行区域，白色区域为不可通行区域。
对于类似图2中的非结构化道路，可以通过一定的处理，从而确定其第一骨架线。此处的处理，是栅格化处理，通过栅格化处理确定第一骨架线的步骤如下：
首先，对非结构化道路中的可通行区域进行栅格化处理，将可通行区域划分为多个栅格；其次，根据每个栅格至不可通行区域的距离，确定每个栅格的栅格值；最后，当任一栅格值表示的为距离不可通行区域中的不同部分时，根据任一栅格值，确定第一骨架线。
下面结合图3，对如何确定第一骨架线进行说明。
图3中黑色区域为不可通行区域，灰色区域为可通行区域，将不可通行区域视为障碍物，将图3中的非结构化道路的边界视为单独的障碍物，将可通行区域进行栅格化处理，每个栅格值表示到最近障碍物的距离，障碍物占据的栅格的栅格值为0，与障碍物相邻的栅格的栅格值记为1，依次向外延伸，直到由两个不同障碍物延伸到相同的栅格结束，比如栅格值为4，表示距离左右两侧的障碍物的距离都为4，则栅格值为4的栅格表示由两个不同障碍物延伸到相同的栅格，此时，延伸终止的栅格组成了该区域的第一骨架线。
图4为图3的区块图，参见图4，根据白线可以将可通行区域分为5个区块，对于1区而言，距离1区的点最近的障碍物是左侧的障碍物，其他侧的 障碍物都没有左侧障碍物近。对于2区而言，距离2区最近的障碍物是上侧的障碍物，其他侧的障碍物都没有上侧的障碍物近。同理，对于5区而言，距离5区最近的障碍物就是中间的障碍物了，其他侧的障碍物都没有中间的障碍物近。
所以说，栅格值4的栅格和顶点的连线就是用来区分不同区块的，栅格值4的栅格和顶点的连线，其上侧的点距离上侧的障碍物更近，其下侧的点距离左侧的障碍物更近，连线恰好是分界线，因此被用来做骨架线。
同理，栅格值4的栅格取中心点的原因也类似，因为中心点距离左侧、上侧、中间障碍物的距离都是一样的，所以作为分界点。
由此，根据图3和图4所示，可以确定图2的非结构化道路的第一骨架线，图2所示的非结构化道路的第一骨架线如图5所示，图5为本发明实施例一提供的分结构化道路的第一骨架线。
其中，上述进行栅格化处理时的栅格大小，是根据车辆的宽度确定的，比如，栅格的大小和车辆的宽度成固定的比例，比如1:1，又或者，栅格的大小是根据车辆的长度和宽度确定的，比如栅格的大小为车辆的长度的平方加上宽度的平方后，再开方，从而根据车辆的长度和宽度，确定栅格大小，便于提高栅格处理时的速度。
步骤120，根据第一骨架线中的每个点与非结构化道路的边界的距离，确定第二骨架线。
具体的，对于第一骨架线，其是由多个边线组成的，当第一骨架线中的边线上的点距离可通行区域的边界的距离小于预设的距离阈值时，将第一骨架线中的边线删除，得到第二骨架线。
比如，第一骨架线上某个边线中如果有一个点距离非结构化道路的边界的距离小于预设距离阈值，则剔除该边线。
步骤130，获取可通行区域的靠右行驶路径。
具体的，在一个示例中，可以根据无人驾驶车辆的宽度和最小转弯半径， 确定靠右行驶时的路径到非结构化道路的边界的距离，从而可以根据该距离和非结构化道路的边界，确定靠右行驶路径。其中，此处的距离，可以是车辆的宽度的一半与最小转弯半径之和。
在另一个示例中，可以获取与不可通行区域相邻的栅格；根据与不可通行区域相邻的栅格的栅格值，确定靠右行驶路径。比如，参见图3，与不可通行区域相邻的栅格值为1，则根据栅格值1，取栅格值为1的栅格的中心点进行连线，从而确定靠右行驶路径。
步骤140，根据第二骨架线和靠右行驶路径，确定非结构化道路的多条边线和多个顶点，多条边线和多个顶点组成类结构化地图；边线具有长度属性和方向属性，任意两条边线的交点为非结构化道路的顶点。
具体的，类结构化地图其实也就是类似结构化地图，在类结构化地图中，非结构化道路被处理为含有多条边线和多个顶点。
参见图6，图6为本发明实施例一提供的第二骨架线及靠右行驶路径示意图，图6中实线为第二骨架线，虚线为靠右行驶路径，图6中的第二骨架线和靠右行驶路径，称为边线，每条边线的交点，称为顶点，两个顶点之间为一条边线，边线不仅具有长度属性，还具有方向属性，边线中的第二骨架线为双向可通行，靠右行驶路径遵循靠右行驶的原则。
下面关于如何确定边线的方向属性进行说明：
具体的，首先获取靠右行驶路径上的任一边线上的任意一点；然后判断任意一点的第一侧点和第二侧点是否在非结构化道路内；第一侧点和第二侧点设置在任意一点的两侧；最后，当第一侧点不在非结构化道路内时，确定任一边线的方向属性为第一方向，当第二侧点不在非结构化道路内时，确定任一边线的方向属性为第二方向；第一方向和第二方向互为反方向。
参见图7，以边线AB为例，如果AB到非结构化道路的边界的距离为l，给定AB一个初始方向如图中箭头所示，即由A指向B，在AB边上任取一点C，沿与C点法线方向在C点左右两侧取点C l和C r。C l和C r符合如下公式：
CC r＝CC r＝l+ε                                   公式(1)
其中，CC r为点C到点C r的线段长度，CC l为点C到点C l的线段长度，ε
Figure PCTCN2021118566-appb-000001
是一个足够小的值，可以看做无穷小。
可以通过检测C l和C r是否在非结构化道路内，确定边线AB的方向属性。如果C r不在非结构化道路内，则图8中箭头即为AB边线的方向，若C l不在非结构化道路内，则图8中箭头的反方向即为AB边线的方向。
参见图8，非结构化道路的类结构化地图中，包括边线的长度属性、方向属性以及多个顶点。该方向属性可以通过箭头进行直观的表示，长度属性可以存储在数据库中，比如，边线具有边线ID，对应的边线ID对应该边线的长度属性。该边线ID和长度属性可以以表格的方式存储在数据库中。如边线AB的ID为01，对应的长度为5m，则将01与5米以表格的方式存储在数据库中。
步骤150，在进行全局路径规划前，将车辆当前位置和目标点映射到类结构化地图上。
具体的，自动驾驶车辆在行驶过程中，可以通过车辆上安装的传感器，比如惯性测量单元(Inertial measurement unit，IMU)、全球定位系统(Global Positioning System，GPS)等得到车辆当前的位置信息。
当服务器接收到用户终端的叫车服务信息时，比如，用户终端接收用户输入的出发地、目的地和出发时间等信息，服务器接收用户终端发送的出发地、目的地信息后，筛选出目标车辆，并将出发地、目的地和出发时间、出发人数等信息一起发送给目标车辆。
对于目标车辆，当车辆行驶过程中，需要进行全局路径规划，以便从出发地行驶至目的地，在进行全局路径规划之前，如果当前位置信息和目的地，即目标点，都处于上述得到的具有多条边线和多个顶点的类结构化地图中，可以将车辆当前位置和目标点映射在类结构化地图中。
其中，车辆当前位置可以是经纬度数据，而类结构化地图中的每个点，都具有位置信息，可以将车辆当前的经纬度数据映射到结构化地图中去，从 而可以确定结构化地图中车辆的当前位置和目标点。
步骤160，根据车辆当前位置在类结构化地图中的第一位置，以及目标点在类结构化地图中的第二位置，确定第一位置到类结构化地图的边线上的第一最近点集，以及第二位置到类结构化地图的边线上的第二最近点集。
具体的，步骤160包括：首先，以第一位置为圆心，以预设第一半径为半径确定起始圆；其次，确定起始圆中的第一最近点集，并将第一最近点集中的每一第一最近点作为全局路径规划的起点；再次，以第二位置为圆心，以预设第二半径为半径确定终点圆；最后，确定终点圆中的第二最近点集，并将第二最近点集中的每一第二最近点作为劝募路径规划的终点。
其中，预设第一半径和预设第二半径可以是相同的，也可以是不同的。预设第一半径可以是根据车辆的宽度和固定值而确定，该固定值与第一位置周围的边线的数量和周围相邻边线的距离相关，比如，参见图9，第一位置为下方的小矩形框，第一位置处有三条边线，则以预设第一半径做圆后，要保证三条边线都处于起始圆中，得到的第一最近点集以下方的小矩形框周围的三个小五角星表示。
相应的，对终点圆的设置与对起始圆的设置方式相似，此处不再赘述。第二位置为上方的小矩形框，第一位置处有三条边线，则以预设第二半径做圆后，得到的第二最近点集以上方的小矩形框周围的三个小五角星表示。
步骤170，根据第一最近点集中的每个第一最近点到采样点沿边线的长度、每个第一最近点到采样点沿边线的方向以及采样点至每个第二最近点的距离，确定第一最近点集中的每个第一最近点至第二最近点集中的每个第二最近点的评价函数值。
具体的，步骤170包括：确定每个起点到采样点沿边线的长度；确定每个起点到采样点沿边线的方向；确定预设的第一权重值和长度的第一乘积；确定预设的第二权重值和方向的第二乘积；确定第一乘积和第二乘积的和值作为第一代价值；确定采样点到目标点的距离作为第二代价值；根据第一代 价值和第二代价值，确定评价函数值。
采样点为在进行每次路径搜索时的一个点，路径搜索可以采用路径搜索算法来实现。比如启发式路径搜索算法中的A*算法等，本申请对此并不限定。
具体的，在一个示例中，当采用启发式路径搜索算法时，可以通过A*算法对应的评价评价函数进行评价，评价函数如公式(2)所示：
f＝g+h                                        公式(2)
g值为搜索的第一代价值，代表从搜索的起始点到当前采样点搜索所付出的代价。h值为当前采样点到目标点的第二代价值，可以用采样点到目标点的距离表示，此处的距离，包括但不限于欧式距离和曼哈顿距离。
其中，本申请可以通过公式(3)计算第一代价值。
g＝α*l+β*dir                                    公式(3)
其中，l表示起始点到当前采样点沿边线的长度，通过顶点将边线进行划分，两个顶点之间的边线作为一条边线，dir为起点到采样点沿边线的方向，其中，顺行记为0，逆行记为1，α为第一权重系数，表示路径长度对整条路的重要性，β为第二权重系数，表示方向对于整条路径的重要性。
α和β可以根据需求进行设定，比如如果不要求车辆正逆行的话，只要求能最短到达目标点的话，那么β的意义就不大，此时可以将β的值设定为比α值小，且β无限小，其具体的数值，需要经过多次模拟计算而确定。
步骤180，确定评价函数值最小的第一最近点至第二最近点之间的路径为最优路径。
后续，可以根据第一代价值和第二代价值，确定评价函数值，并根据评价函数值，确定评价函数值最小的对应最优路径。
比如，起始点有1、2、3，采样点为A，终点为4,5,6，则起始点至采样点，一共计算1-A、2-A、3-A三条路径的第一代价值，采样点至终点，计算A-4、A-5和A-6三条路径的第二代价值，然后任意相加，即1-A+A-4，1-A+A-5，1-A+A-6，2-A+A-4，2-A+A-5，2-A+A-6，3-A+A-4，3-A+A-5，3-A+A-6， 计算这9个值，确定评价函数值最小的为一个局部路径，比如1-A+A-4最小，可以确定起始点1到采样点A再到终点4为最优路径。
通过应用本发明实施例提供的全局路径规划方法，可以将非结构化道路进行拓扑化处理，得到其类结构化地图，从而通过类结构化地图经全局路径规划，以提高全局路径规划的效率，同时，本申请在进行路径评价时，可以结合类结构化地图中的边线的长度属性和方向属性，从而提升了路径搜索结果的合理性。
图10为本发明实施例二提供的全局路径规划装置结构示意图，如图10所示，该全局路径规划装置应用在实施例一中的全局路径规划方法中，全局路径规划装置包括：获取单元210、确定单元220和映射单元230。
获取单元210用于获取非结构化道路中的可通行区域的第一骨架线；非结构化道路包括可通行区域和不可通行区域；
确定单元220用于根据第一骨架线中的每个点与非结构化道路的边界的距离，确定第二骨架线；
获取单元210还用于，获取可通行区域的靠右行驶路径；
确定单元220还用于，根据第二骨架线和靠右行驶路径，确定非结构化道路的多条边线和多个顶点，多条边线和多个顶点组成类结构化地图；边线具有长度属性和方向属性，任意两条边线的交点为非结构化道路的顶点；
映射单元230用于在进行全局路径规划前，将车辆当前位置和目标点映射到类结构化地图上；
确定单元220还用于，根据车辆当前位置在类结构化地图中的第一位置，以及目标点在类结构化地图中的第二位置，确定第一位置到类结构化地图的边线上的第一最近点集，以及第二位置到类结构化地图的边线上的第二最近点集；
确定单元220还用于，根据第一最近点集中的每个第一最近点到采样点沿边线的长度、每个第一最近点到采样点沿边线的方向以及采样点至每个第 二最近点的距离，确定第一最近点集中的每个第一最近点至第二最近点集中的每个第二最近点的评价函数值；
确定单元220还用于，确定评价函数值最小的第一最近点至第二最近点之间的路径为最优路径。
进一步的，获取单元210具体用于：对非结构化道路中的可通行区域进行栅格化处理，将可通行区域划分为多个栅格；根据每个栅格至不可通行区域的距离，确定每个栅格的栅格值；当任一栅格值表示的为距离不可通行区域中的不同部分时，根据任一栅格值，确定第一骨架线。
进一步的，获取单元210具体用于：获取与不可通行区域相邻的栅格；根据与不可通行区域相邻的栅格的栅格值，确定靠右行驶路径。
其中，进行栅格化处理时，根据车辆的宽度确定栅格大小；根据车辆的宽度和最小转弯半径，确定靠右行驶路径到非结构化道路的边界的距离。
进一步的，确定单元220具体用于：
当第一骨架线中的点距离可通行区域的边界的距离小于预设的距离阈值时，将第一骨架线中的点删除，得到第二骨架线。
其中，边线具有方向属性具体包括：
获取靠右行驶路径上的任一边线上的任意一点；
判断任意一点的第一侧点和第二侧点是否在非结构化道路内；第一侧点和第二侧点设置在任意一点的两侧；
当第一侧点不在非结构化道路内时，确定任一边线的方向属性为第一方向，当第二侧点不在非结构化道路内时，确定任一边线的方向属性为第二方向；第一方向和第二方向互为反方向。
进一步的，确定单元220具体用于：
以第一位置为圆心，以预设第一半径为半径确定起始圆；
确定起始圆中的第一最近点集，并将第一最近点集中的每一第一最近点作为全局路径规划的起点；
以第二位置为圆心，以预设第二半径为半径确定终点圆；
确定终点圆中的第二最近点集，并将第二最近点集中的每一第二最近点作为劝募路径规划的终点。
进一步的，确定单元220具体用于：
确定每个起点到采样点沿边线的长度；确定每个起点到采样点沿边线的方向；确定预设的第一权重值和长度的第一乘积；确定预设的第二权重值和方向的第二乘积；确定第一乘积和第二乘积的和值作为第一代价值；确定采样点到目标点的距离作为第二代价值；根据第一代价值和第二代价值，确定评价函数值。
通过应用本发明实施例提供的全局路径规划装置，可以将非结构化道路进行拓扑化处理，得到其类结构化地图，从而通过类结构化地图经全局路径规划，以提高全局路径规划的效率，同时，本申请在进行路径评价时，可以结合类结构化地图中的边线的长度属性和方向属性，从而提升了路径搜索结果的合理性。
本发明实施例三提供了一种设备，包括存储器和处理器，存储器用于存储程序，存储器可通过总线与处理器连接。存储器可以是非易失存储器，例如硬盘驱动器和闪存，存储器中存储有软件程序和设备驱动程序。软件程序能够执行本发明实施例提供的上述方法的各种功能；设备驱动程序可以是网络和接口驱动程序。处理器用于执行软件程序，该软件程序被执行时，能够实现本发明实施例一提供的方法。
本发明实施例四提供了一种包含指令的计算机程序产品，当计算机程序产品在计算机上运行时，使得计算机执行本发明实施例一提供的方法。
本发明实施例五提供了一种计算机可读存储介质，计算机可读存储介质上存储有计算机程序，计算机程序被处理器执行时实现本发明实施例一提供的方法。
本发明实施例六提供了一种计算机服务器，包括：存储器、处理器和收 发器；所述处理器用于与所述存储器耦合，读取并执行所述存储器中的指令，以实现如实施例一所述的全局路径规划方法；所述收发器与所述处理器耦合，由所述处理器控制所述收发器进行消息收发。
本发明实施例七提供了一种移动工具，包括如实施例六的计算机服务器。上述移动工具可以是任何可以移动的工具，例如车辆(例如吸尘车、清扫车、洗地车、物流小车、乘用车、环卫车、公交车、大巴车、厢式货车、卡车、载重车、挂车、甩挂车、吊车、挖掘机、铲土机、公路列车、扫地车、洒水车、垃圾车、工程车、救援车、AGV(Automated Guided Vehicle，自动导引运输车)、摩托车、自行车、三轮车、手推车、机器人、扫地机、平衡车等，本申请对于移动工具的类型不做严格限定，在此不再穷举。专业人员应该还可以进一步意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、计算机软件或者二者的结合来实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
结合本文中所公开的实施例描述的方法或算法的步骤可以用硬件、处理器执行的软件模块，或者二者的结合来实施。软件模块可以置于随机存储器(RAM)、内存、只读存储器(ROM)、电可编程ROM、电可擦除可编程ROM、寄存器、硬盘、可移动磁盘、CD-ROM、或技术领域内所公知的任意其它形式的存储介质中。
以上的具体实施方式，对本发明的目的、技术方案和有益效果进行了进一步详细说明，所应理解的是，以上仅为本发明的具体实施方式而已，并不用于限定本发明的保护范围，凡在本发明的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (12)

  1. 一种全局路径规划方法，其特征在于，包括：
    获取非结构化道路中的可通行区域的第一骨架线；所述非结构化道路包括可通行区域和不可通行区域；
    根据所述第一骨架线中的每个点与所述非结构化道路的边界的距离，确定第二骨架线；
    获取所述可通行区域的靠右行驶路径；
    根据所述第二骨架线和所述靠右行驶路径，确定所述非结构化道路的多条边线和多个顶点，所述多条边线和所述多个顶点组成类结构化地图；所述边线具有长度属性和方向属性，任意两条边线的交点为所述非结构化道路的顶点；
    在进行全局路径规划前，将车辆当前位置和目标点映射到所述类结构化地图上；
    根据所述车辆当前位置在所述类结构化地图中的第一位置，以及所述目标点在所述类结构化地图中的第二位置，确定所述第一位置到所述类结构化地图的边线上的第一最近点集，以及所述第二位置到所述类结构化地图的边线上的第二最近点集；
    根据所述第一最近点集中的每个第一最近点到采样点沿边线的长度、每个第一最近点到采样点沿边线的方向以及采样点至每个第二最近点的距离，确定所述第一最近点集中的每个第一最近点至所述第二最近点集中的每个第二最近点的评价函数值；
    确定评价函数值最小的第一最近点至第二最近点之间的路径为最优路径。
  2. 根据权利要求1所述的方法，其特征在于，所述获取非结构化道路中的可通行区域的第一骨架线具体包括:
    对所述非结构化道路中的可通行区域进行栅格化处理，将所述可通行区域划分为多个栅格；
    根据每个栅格至所述不可通行区域的距离，确定每个栅格的栅格值；
    当任一栅格值表示的为距离所述不可通行区域中的不同部分时，根据所述任一栅格值，确定第一骨架线。
  3. 根据权利要求2所述的方法，其特征在于，所述获取所述可通行区域的靠右行驶路径具体包括：
    获取与所述不可通行区域相邻的栅格；
    根据与所述不可通行区域相邻的栅格的栅格值，确定靠右行驶路径。
  4. 根据权利要求2或3所述的方法，其特征在于，进行栅格化处理时，根据车辆的宽度确定栅格大小；
    根据车辆的宽度和最小转弯半径，确定所述靠右行驶路径到所述非结构化道路的边界的距离。
  5. 根据权利要求1-4中任意一项所述的方法，其特征在于，所述根据所述第一骨架线中的每个点与所述非结构化道路的边界的距离，确定第二骨架线具体包括：
    当所述第一骨架线中的点距离所述可通行区域的边界的距离小于预设的距离阈值时，将所述第一骨架线中的点删除，得到第二骨架线。
  6. 根据权利要求1-5中任意一项所述的方法，其特征在于，所述边线具有方向属性具体包括：
    获取靠右行驶路径上的任一边线上的任意一点；
    判断所述任意一点的第一侧点和第二侧点是否在非结构化道路内；所述第一侧点和所述第二侧点设置在所述任意一点的两侧；
    当所述第一侧点不在所述非结构化道路内时，确定所述任一边线的方向属性为第一方向，当所述第二侧点不在所述非结构化道路内时，确定所述任一边线的方向属性为第二方向；所述第一方向和所述第二方向互为反方向。
  7. 根据权利要求1-6中任意一项所述的方法，其特征在于，所述搜索所述第一位置到类结构化地图的边线上的第一最近点集，以及所述第二位置到 类结构化地图的边线上的第二最近点集具体包括：
    以所述第一位置为圆心，以预设第一半径为半径确定起始圆；
    确定所述起始圆中的第一最近点集，并将所述第一最近点集中的每一第一最近点作为全局路径规划的起点；
    以所述第二位置为圆心，以预设第二半径为半径确定终点圆；
    确定所述终点圆中的第二最近点集，并将所述第二最近点集中的每一第二最近点作为劝募路径规划的终点。
  8. 根据权利要求7所述的方法，其特征在于，所述根据所述第一最近点集中的每个第一最近点到采样点沿边线的长度、每个第一最近点到采样点沿边线的方向以及采样点至每个第二最近点的距离，确定第一最近点集中的每个第一最近点至第二最近点集中的每个第二最近点的评价值具体包括：
    确定每个起点到采样点沿边线的长度；
    确定每个起点到采样点沿边线的方向；
    确定预设的第一权重值和所述长度的第一乘积；
    确定预设的第二权重值和所述方向的第二乘积；
    确定所述第一乘积和所述第二乘积的和值作为第一代价值；
    确定所述采样点到目标点的距离作为第二代价值；
    根据所述第一代价值和所述第二代价值，确定评价函数值。
  9. 一种全局路径规划装置，其特征在于，所述全局路径规划装置包括：
    获取单元，所述获取单元用于获取非结构化道路中的可通行区域的第一骨架线；所述非结构化道路包括可通行区域和不可通行区域；
    确定单元，所述确定单元用于根据所述第一骨架线中的每个点与所述非结构化道路的边界的距离，确定第二骨架线；
    所述获取单元还用于，获取所述可通行区域的靠右行驶路径；
    所述确定单元还用于，根据所述第二骨架线和所述靠右行驶路径，确定非结构化道路的多条边线和多个顶点，所述多条边线和所述多个顶点组成类 结构化地图；所述边线具有长度属性和方向属性，任意两条边线的交点为非结构化道路的顶点；
    映射单元，所述映射单元用于在进行全局路径规划前，将车辆当前位置和目标点映射到所述类结构化地图上；
    所述确定单元还用于，根据车辆当前位置在所述类结构化地图中的第一位置，以及目标点在所述类结构化地图中的第二位置，确定所述第一位置到类结构化地图的边线上的第一最近点集，以及所述第二位置到类结构化地图的边线上的第二最近点集；
    所述确定单元还用于，根据所述第一最近点集中的每个第一最近点到采样点沿边线的长度、每个第一最近点到采样点沿边线的方向以及采样点至每个第二最近点的距离，确定第一最近点集中的每个第一最近点至第二最近点集中的每个第二最近点的评价函数值；
    所述确定单元还用于，确定评价函数值最小的第一最近点至第二最近点之间的路径为最优路径。
  10. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质上存储有计算机程序，计算机程序被处理器执行时执行如权利要求1-8任一所述的方法。
  11. 一种计算机服务器，其特征在于，包括：存储器、处理器和收发器；
    所述处理器用于与所述存储器耦合，读取并执行所述存储器中的指令，以实现权利要求1-8任一项所述的全局路径规划方法；
    所述收发器与所述处理器耦合，由所述处理器控制所述收发器进行消息收发。
  12. 一种移动工具，其特征在于，包括如权利要求11所述的计算机服务器。
PCT/CN2021/118566 2020-09-22 2021-09-15 全局路径规划方法及装置 WO2022063005A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011003654.1A CN112444263B (zh) 2020-09-22 2020-09-22 全局路径规划方法及装置
CN202011003654.1 2020-09-22

Publications (1)

Publication Number Publication Date
WO2022063005A1 true WO2022063005A1 (zh) 2022-03-31

Family

ID=74736831

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/118566 WO2022063005A1 (zh) 2020-09-22 2021-09-15 全局路径规划方法及装置

Country Status (2)

Country Link
CN (1) CN112444263B (zh)
WO (1) WO2022063005A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116431964A (zh) * 2023-04-20 2023-07-14 浙江省水利河口研究院(浙江省海洋规划设计研究院) 一种复杂河网水系骨架线生成的游程剥离法
CN116543310A (zh) * 2023-06-30 2023-08-04 眉山环天智慧科技有限公司 一种基于Voronoi图和核密度的道路线提取方法

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112444263B (zh) * 2020-09-22 2023-05-23 北京智行者科技股份有限公司 全局路径规划方法及装置
CN113119995B (zh) * 2021-03-11 2023-01-31 京东鲲鹏(江苏)科技有限公司 一种路径搜索方法及装置、设备、存储介质
CN113219973B (zh) * 2021-05-08 2022-06-24 浙江工业大学 一种移动机器人的局部路径控制方法
CN113670323A (zh) * 2021-08-17 2021-11-19 京东鲲鹏(江苏)科技有限公司 一种目标区域的确定方法、装置、设备和介质
CN113928337B (zh) * 2021-09-30 2023-10-20 华为技术有限公司 引导车辆行驶的方法及相关系统、存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050234638A1 (en) * 2004-04-05 2005-10-20 Tadao Ogaki Navigation apparatus, and data processing method and computer program used therewith
CN106767860A (zh) * 2016-11-21 2017-05-31 江苏大学 一种基于启发式搜索算法来缩短智能汽车路径规划搜索时间的方法
CN107664503A (zh) * 2016-07-29 2018-02-06 上海汽车集团股份有限公司 车辆路径规划方法及装置
CN107992050A (zh) * 2017-12-20 2018-05-04 广州汽车集团股份有限公司 无人驾驶汽车局部路径运动规划方法和装置
CN111521189A (zh) * 2020-04-10 2020-08-11 北京智行者科技有限公司 清扫路径规划方法及装置
CN112444263A (zh) * 2020-09-22 2021-03-05 重庆智行者信息科技有限公司 全局路径规划方法及装置

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9524647B2 (en) * 2015-01-19 2016-12-20 The Aerospace Corporation Autonomous Nap-Of-the-Earth (ANOE) flight path planning for manned and unmanned rotorcraft
CN106092102A (zh) * 2016-07-20 2016-11-09 广州极飞电子科技有限公司 一种无人机路径规划方法及装置
CN109557928A (zh) * 2019-01-17 2019-04-02 湖北亿咖通科技有限公司 基于矢量地图和栅格地图的自动驾驶车辆路径规划方法
CN110850871B (zh) * 2019-10-21 2020-07-17 深圳市银星智能科技股份有限公司 一种机器路径规划方法及移动机器人
CN110823241B (zh) * 2019-11-19 2021-05-28 齐鲁工业大学 基于可通行区域骨架提取的机器人路径规划方法及系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050234638A1 (en) * 2004-04-05 2005-10-20 Tadao Ogaki Navigation apparatus, and data processing method and computer program used therewith
CN107664503A (zh) * 2016-07-29 2018-02-06 上海汽车集团股份有限公司 车辆路径规划方法及装置
CN106767860A (zh) * 2016-11-21 2017-05-31 江苏大学 一种基于启发式搜索算法来缩短智能汽车路径规划搜索时间的方法
CN107992050A (zh) * 2017-12-20 2018-05-04 广州汽车集团股份有限公司 无人驾驶汽车局部路径运动规划方法和装置
CN111521189A (zh) * 2020-04-10 2020-08-11 北京智行者科技有限公司 清扫路径规划方法及装置
CN112444263A (zh) * 2020-09-22 2021-03-05 重庆智行者信息科技有限公司 全局路径规划方法及装置

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116431964A (zh) * 2023-04-20 2023-07-14 浙江省水利河口研究院(浙江省海洋规划设计研究院) 一种复杂河网水系骨架线生成的游程剥离法
CN116431964B (zh) * 2023-04-20 2024-04-19 浙江省水利河口研究院(浙江省海洋规划设计研究院) 一种复杂河网水系骨架线生成的游程剥离法
CN116543310A (zh) * 2023-06-30 2023-08-04 眉山环天智慧科技有限公司 一种基于Voronoi图和核密度的道路线提取方法
CN116543310B (zh) * 2023-06-30 2023-10-31 眉山环天智慧科技有限公司 一种基于Voronoi图和核密度的道路线提取方法

Also Published As

Publication number Publication date
CN112444263B (zh) 2023-05-23
CN112444263A (zh) 2021-03-05

Similar Documents

Publication Publication Date Title
WO2022063005A1 (zh) 全局路径规划方法及装置
CN111426330B (zh) 路径生成方法和设备、无人化运输系统和存储介质
US11126187B2 (en) Systems and methods for controlling the operation of a vehicle
EP3517893B1 (en) Path and speed optimization fallback mechanism for autonomous vehicles
CN109263639B (zh) 基于状态栅格法的驾驶路径规划方法
González et al. Continuous curvature planning with obstacle avoidance capabilities in urban scenarios
US10095234B2 (en) Planning for unknown objects by an autonomous vehicle
US20190232957A1 (en) Planning for unknown objects by an autonomous vehicle
WO2019042295A1 (zh) 一种无人驾驶路径规划方法、系统和装置
CN114222691A (zh) 基于成本的路径确定
CN108088456A (zh) 一种具有时间一致性的无人驾驶车辆局部路径规划方法
CN113916246A (zh) 一种无人驾驶避障路径规划方法和系统
US20180259967A1 (en) Planning for unknown objects by an autonomous vehicle
JP2021046193A (ja) 混雑した道路における協働認識車線変更制御を提供するためのシステム及び方法
CN114258366B (zh) 对于自主车辆的折线轮廓表示
CN112284393B (zh) 一种智能移动机器人全局路径规划方法和系统
CN104897168A (zh) 基于道路危险评估的智能车路径搜索方法及系统
US11834077B2 (en) Systems, methods, and media for occlusion-aware motion planning
CN113619603B (zh) 一种双阶段自动驾驶车辆调头轨迹规划方法
CN113494923A (zh) 基于微分动态规划的路径轨迹规划方法及系统
CN116723970A (zh) 使用界限表示进行自动驾驶拓扑规划的系统、方法和计算机程序产品
US11584371B2 (en) Systems and methods for using R-functions and semi-analytic geometry for lane keeping in trajectory planning
CN115221774A (zh) 自主车辆交通模拟和道路网络建模
CN112149484A (zh) 基于车道线确定消失点
US11480962B1 (en) Dynamic lane expansion

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21871376

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21871376

Country of ref document: EP

Kind code of ref document: A1