Unmanned vehicle path planning method and system suitable for complex environment

Abstract

The invention provides an unmanned vehicle path planning method and system suitable for a complex environment, and relates to the technical field of intelligent vehicle navigation and control, and the method comprises the steps: obtaining a basic grid map of a field complex environment; calculating the grid gradient of each grid according to the basic grid map; setting different earth surface attributes according to different earth surface information in the basic grid map and the passing degree of the unmanned vehicle on different earth surface information; overlapping the grid gradient and the earth surface attribute into an elevation information layer of a corresponding grid in the basic grid map to obtain a grid overlapping map; improving the A * algorithm by using a neighborhood search method to obtain an improved A * algorithm; and determining a starting grid and a target grid in the grid superposition map, and performing path planning on the grid superposition map by using an improved A * algorithm to obtain an optimal path from the starting grid to the target grid. According to the method, the problem that various algorithms in the prior art are difficult to meet unmanned vehicle path planning in a field complex environment is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent vehicle navigation and control, and particularly to a path planning method and system for an unmanned vehicle suitable for complex environments. Background Art

[0002] With the rapid development of driverless technology, its applications have gradually expanded to various fields such as urban transportation and logistics, especially showing higher safety and flexibility in fields such as hazardous transportation, emergency rescue, and environmental reconnaissance. However, current research on unmanned vehicle path planning mainly focuses on simple structured environments, and the research on complex field environments is relatively insufficient. Field environments are usually highly complex, including variable terrains, dynamic obstacles, and unpredictable weather conditions, etc., which pose severe challenges to the path planning of unmanned vehicles. And path planning is not only the core technology of unmanned vehicle autonomous navigation but also the basis for ensuring the efficient and safe movement of unmanned vehicles. Therefore, it is urgent to conduct in-depth research on the path planning of unmanned vehicles in complex field environments to improve the navigation ability and application effect of unmanned vehicles in complex environments.

[0003] At present, according to the degree of mastery of environmental state information, the path planning algorithms for autonomous vehicles can be mainly divided into global path planning algorithms and local path planning algorithms. Global path planning algorithms include Dijkstra's algorithm, A* algorithm, Rapidly-exploring Random Tree (RRT), Probabilistic Roadmap (PRM), DQN algorithm, etc.; local path planning algorithms include Dynamic Window Approach (DWA), Artificial Potential Field (APF), etc. Each algorithm has its own advantages and disadvantages. Dijkstra's algorithm is a classic shortest path algorithm that can guarantee finding the optimal path from the starting point to the ending point. However, it has many expanded nodes and low computational efficiency, and it cannot effectively adjust the real-time path in a dynamic environment. Although the A* algorithm performs well in a fixed and known environment, due to its insufficient adaptability to environmental changes, it often fails to meet the requirements of real-time and flexibility in complex outdoor environments. The RRT algorithm can quickly explore paths in high-dimensional spaces, is suitable for complex unstructured environments, and can effectively handle dynamic constraints. However, its path is not smooth enough and it is prone to falling into local optima. PRM constructs a probability graph and can effectively handle path planning problems in high-dimensional configuration spaces. However, this algorithm is only suitable for static environments. When facing complex environments, the effectiveness and efficiency of the algorithm will be limited by the sparsity of the complex environment, and frequent re-sampling and reconstruction of the road network are required in dynamic complex environments, affecting real-time and flexibility. DWA is a real-time responsive local path planning method that can quickly generate feasible paths in a dynamic environment by considering the motion model of the autonomous vehicle and environmental information. However, its algorithm mainly relies on local information in the current state to make decisions, and its path depends on speed and acceleration limits, resulting in an insufficiently smooth path. Although APF has a small amount of computation and can avoid obstacles in real time, it is prone to falling into local minima, which may lead to the inability to find an effective path and may not be able to effectively handle multiple obstacles in complex environments.

[0004] Regarding various existing algorithms and their improvements, although they each have their own characteristics in specific environments, they still have many limitations and are difficult to meet the path planning of unmanned vehicles in complex outdoor environments. On the one hand, existing research usually only focuses on basic factors such as path length, obstacle avoidance ability, and travel time, and mostly relies on traditional grid maps for path planning. However, traditional grid maps have significant limitations in information expression. Especially when dealing with complex dynamic environments, it is difficult to effectively adapt to changing terrains and dynamic obstacles. On the other hand, the traditional A* algorithm has a fast calculation speed and is one of the most effective and common search methods for solving the shortest path. Its search mechanism is usually limited to four-way search or eight-way search, which causes two main problems in path planning: one is that due to the limitation of the search range, the generated path often shows non-smooth characteristics, resulting in unnecessary sharp turns or non-stable driving of the unmanned vehicle during actual operation; the other is that this limitation makes the algorithm more likely to fall into local optimal solutions, especially difficult to find the global optimal path in complex and dynamically changing environments. Summary of the Invention

[0005] In view of the above problems, the present invention provides a path planning method and system for unmanned vehicles applicable to complex environments to solve the problem that various existing algorithms are difficult to meet the path planning of unmanned vehicles in complex outdoor environments.

[0006] On the one hand, the present invention provides a path planning method for unmanned vehicles applicable to complex environments, and the method includes:

[0007] Obtain the basic grid map of the complex outdoor environment;

[0008] Calculate the grid slope of each grid according to the basic grid map; set different surface attributes according to different surface information in the basic grid map and the passability of the unmanned vehicle on different surface information; superimpose the grid slope and the surface attributes into the elevation information layer of the corresponding grid in the basic grid map to obtain a grid superimposed map;

[0009] Improve the A* algorithm using the neighborhood search method to obtain an improved A* algorithm;

[0010] Determine the starting grid and the destination grid in the grid superimposed map, and use the improved A* algorithm to perform path planning on the grid superimposed map to obtain the optimal path from the starting grid to the destination grid.

[0011] Further, setting different surface attributes includes:

[0012] Match different color codes and / or texture codes for each grid in the basic grid map according to the passability corresponding to different surface information.

[0013] Furthermore, the basic attributes of the raster overlay map include: raster abscissa, raster ordinate, elevation information, raster slope, and surface attributes.

[0014] Furthermore, the calculation method of the raster slope includes: regarding a small piece of the surface curve corresponding to each raster as a plane, and the dihedral angle value between the plane and the horizontal plane is the raster slope.

[0015] Furthermore, the calculation formula for the raster slope s is:

[0016]

[0017] where represents the slope change rate of the current raster in the x direction of the raster abscissa; represents the slope change rate of the current raster in the y direction of the raster ordinate.

[0018] Furthermore, improving the A* algorithm using the neighborhood search method includes:

[0019] Expanding the search neighborhood to multiple layers of rasters centered on the current raster, so that the path search can not only search the first layer of rasters adjacent to the current raster, but also search the second layer of rasters adjacent to the first layer of rasters.

[0020] Furthermore, when performing neighborhood search, the conditions for generating neighborhood rasters need to satisfy both the first condition and the second condition; where

[0021] The first condition is: the neighborhood raster is not occupied by an obstacle or is a passable area;

[0022] The second condition is: the neighborhood raster is within the range of the raster overlay map.

[0023] Furthermore, for different complex field environments, one or more priority heuristic functions based on timeliness, safety, and smoothness are constructed, and the improved A* algorithm is used to perform path planning on the raster overlay map.

[0024] Furthermore, the first priority heuristic function f t (n) is:

[0025] f t (n) = g t (n) + h t (n);

[0026]

[0027] where time(i) represents the time required to travel on the raster; d nRepresents the distance between the current grid n and the destination grid; v max Represents the maximum speed of the driverless vehicle;

[0028] The second-priority heuristic function f of the safety r (n) is:

[0029] f r (n) = g r (n) + h r (n);

[0030]

[0031] h r (n) = d n ;

[0032] Among them, risk(i) represents the risk time of driving in the grid; d n Represents the distance between the current grid n and the destination grid;

[0033] The third-priority heuristic function f of the smoothness s (n) is:

[0034] f s (n) = g s (n) + h s (n);

[0035]

[0036] h s (n) = d n ·avg;

[0037]

[0038] Among them, s l Represents the slope difference between two adjacent grids in the driving path; avg represents the average slope difference of the driving distance of the driverless vehicle.

[0039] On the other hand, the present invention provides a driverless vehicle path planning system applicable to a complex environment, including a memory, a processor, and a computer program stored on the memory, and the processor executes the computer program to implement the steps of the method described in any one of the above.

[0040] Generally speaking, the present invention provides a driverless vehicle path planning method and system applicable to a complex environment. Through the technical solution conceived by the present invention, the following beneficial effects can be achieved compared with the prior art:

[0041] (1) The present invention calculates the grid slope of each grid based on the basic grid map; at the same time, different surface attributes are set according to different surface information in the basic grid map and the passability of the unmanned vehicle on different surface information, and the grid slope and surface attributes are superimposed on the elevation information layer of the corresponding grid in the basic grid map; by superimposing the elevation information, grid slope and surface attributes, that is, superimposing the ground conditions and potential risk areas in its real environment on each grid, the comprehensive perception of the complex dynamic environment in the wild is enhanced, and then the positions of obstacles and risk areas are accurately described to ensure safe obstacle avoidance; in this way, the obstacle information is combined with the environmental characteristics to construct a multi-level path planning map, which can effectively adapt to the changing terrain environment and dynamic obstacles, improve the accuracy of path planning, be more suitable for the complex environment in the wild, and lay a foundation for providing a more intelligent and efficient path planning for the unmanned vehicle.

[0042] (2) The present invention improves the A* algorithm by using the neighborhood search method, expanding the four-way or eight-way search of the A* algorithm into a multi-level neighborhood search mechanism, which can not only explore possible paths more efficiently, greatly improve the coverage of the search range, reduce the path planning time, but also make the path smoother, avoid unnecessary sharp turns or non-steady driving, make the planned path safer and more stable, and also more meet the requirements of complex terrain.

[0043] (3) The present invention constructs an optimized heuristic function based on timeliness, safety and stability, and processes and optimizes obstacles and paths by expanding the search mechanism of the A* algorithm and establishing an optimized heuristic function, improving the search efficiency and path smoothness in the complex terrain in the wild; it can not only provide different path planning schemes for different environments or tasks, with strong personalized planning, but also be applicable to various types of unmanned vehicles, and can be applied to different scenarios, such as urban traffic, rural roads, complex terrain, etc., with good versatility and adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0045] Figure 1 It is a schematic diagram of the method steps of a method and system for path planning of an unmanned vehicle suitable for a complex environment provided by the present invention;

[0046] Figure 2 It is a basic grid map of a method and system for path planning of an unmanned vehicle suitable for a complex environment provided by the present invention;

[0047] Figure 3 It is the grid overlay map of a method and system for path planning of an unmanned vehicle applicable to complex environments provided by the present invention;

[0048] Figure 4 It is the schematic diagram of slope calculation of a method and system for path planning of an unmanned vehicle applicable to complex environments provided by the present invention;

[0049] Figure 5 It is the schematic diagram of domain search of a method and system for path planning of an unmanned vehicle applicable to complex environments provided by the present invention. Detailed implementation manners

[0050] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings and embodiments in the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] It should be noted that in the description of the embodiments of the present invention, the terms "include", "comprise" or any other variant thereof are intended to cover a non-exclusive inclusion, so that a method, step or device including a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such method, step or device. Without further limitations, an element defined by the phrase "including one..." does not exclude the existence of additional identical elements in the method, step or device including the element.

[0052] To solve the problem that various existing algorithms are difficult to meet the path planning of unmanned vehicles in the wild and complex environments, the present invention provides a method for path planning of an unmanned vehicle applicable to complex environments, as Figure 1 shown. Specifically, the method includes:

[0053] Step 101: Obtain the basic grid map of the wild and complex environment.

[0054] In the research and application of unmanned vehicles, the establishment of an environmental map is a crucial link. Especially in the wild and complex environments, a high-quality environmental map is not only the basis for path planning but also the prerequisite for realizing safe and efficient navigation. The wild and complex environments usually have variable terrains, dynamic obstacles and potential risk factors. Due to the existence of these characteristics, higher requirements are needed for the efficient path planning of unmanned vehicles.

[0055] The grid map is an effective method for representing the environmental map, which is achieved by dividing the environment into uniform grid cells, and each grid cell contains various information about the environment.

[0056] It should be noted that the basic grid map can construct grid maps with different precisions according to the needs of different scenarios. Currently, the basic attributes mainly included in the basic grid map are: grid abscissa, grid ordinate, and elevation information; the basic grid map can be represented as G 0 = [x, y, h]; where x represents the grid abscissa, y represents the grid ordinate, and h represents the elevation information. Information such as obstacles, risk areas, and passable areas is described through these basic attributes.

[0057] As Figure 2 shown, the basic grid map includes obstacles, risk areas, and passable areas. The "black grid" in the basic grid map represents obstacles that the unmanned vehicle cannot pass through in the map, "×" represents that there may be risks when the unmanned vehicle travels in this area, and the "white grid" represents that the unmanned vehicle can pass through in this area.

[0058] Since most of the obstacles in the real environment are irregular in shape and cannot fill the entire grid. Therefore, as an embodiment, the present invention performs dilation processing on the grid including obstacles. That is, as long as the grid involves obstacles, it is regarded as an obstacle area, thereby improving the accuracy of path planning.

[0059] Due to the risks of geological disasters such as landslides and debris flows in the real environment, these disasters will damage the ground, and then situations such as collapses, cracks, and accumulations will occur, which will seriously affect the vehicle's passage and driving safety. Therefore, when the vehicle travels in this area, it is regarded as having risks.

[0060] Although the basic grid map is convenient for the unmanned vehicle to perform effective path planning in a complex environment, however, in the face of risky areas such as dirt roads, grasslands, sandy lands, and woods, the basic grid map is uniformly marked as "×", which greatly reduces the passable area and may make the path from the starting point to the ending point detour a large circle. Although the safety is relatively high, it is time-consuming and laborious, resulting in the inability to plan the optimal route that can balance timeliness, safety, and smoothness in the wild complex environment. Therefore, the present invention establishes a grid overlay map, providing a good foundation for subsequent dynamic information updates.

[0061] Step 102: Calculate the grid slope of each grid according to the basic grid map; set different surface attributes according to different surface information in the basic grid map and the passability of the unmanned vehicle on different surface information; overlay the grid slope and the surface attributes into the elevation information layer of the corresponding grid in the basic grid map to obtain the grid overlay map.

[0062] Since the content information in the basic grid map is scarce, in the complex outdoor environment, different outdoor surface attributes and the passability of the vehicle on different surface attributes have a significant impact on path planning.

[0063] Therefore, the present invention analyzes the ground features in the basic grid map, quantifies the passability of different types of areas through the driving ability of the vehicle in the complex outdoor environment, sets different surface attributes by using the surface information and passability, enhances the comprehensive perception of the complex dynamic outdoor environment, and then accurately describes the positions of obstacles and risk areas; projects the surface attributes and grid slopes onto the elevation information layer, realizes the superposition of information, and obtains a multi-level path planning map, so as to effectively adapt to the changing terrain environment and dynamic obstacles and improve the accuracy of path planning.

[0064] As an embodiment, setting different surface attributes includes: matching different color codes and / or texture codes for each grid in the basic grid map according to different surface information and corresponding passability.

[0065] More specifically, obtain the obstacle areas, passable areas, and dangerous areas in the basic grid map; mark the obstacle areas with the first color code and / or the first texture code; analyze the ground features of the passable areas and / or dangerous areas to obtain the surface information of each grid; and use different ground features and the analysis of the vehicle's passing process on different grounds to obtain the passability of the corresponding grid; finally, set different surface attributes for each grid in the basic grid map according to different surface information and corresponding passability.

[0066] To facilitate the clear distinction of different surface attributes, different surface attributes are set according to different surface information and the passability of the unmanned vehicle on different surface information, and then different colors are matched for each grid. For example, as shown in Table 1, for different surface information in the complex outdoor environment, where the water system area is regarded as an obstacle, that is, the passability is 0.

[0067] Table 1: Surface Information

[0068] Surface information Feature Trafficability Color Hard pavement The surface is relatively flat, and it is easier for vehicles to maintain stability during driving 1 Grey Earth road surface The surface has unevenness, water accumulation and obstacles, etc., and driving is unstable 0.8 Yellow Grassland The ground is flat and well covered with vegetation, and there may be small obstacles 0.6 Light green Sandy land Soft and slippery, and deep sand areas may make it difficult for vehicles to drive 0.4 Pink Woods The trees are dense, the ground is covered with fallen leaves and weeds, and the field of vision is limited 0.2 Dark green Water system Including rivers, lakes and wetlands, vehicles cannot pass through 0 Blue

[0069] Specifically, the surface of the hard pavement is relatively flat, making it easier for vehicles to maintain stability during driving. The trafficability of the vehicle is recorded as 1, and the color can be matched to gray; the surface of the soil pavement has unevenness, water accumulation, obstacles, etc., resulting in unstable driving. The trafficability of the vehicle is recorded as 0.8, and the color can be matched to yellow; the ground of the grassland is flat and has good vegetation coverage, and there may be small obstacles. The trafficability of the vehicle is recorded as 0.6, and the color can be matched to light green; the sand is soft and slippery, and deep sand areas may cause difficulties for vehicles to drive. The trafficability of the vehicle is recorded as 0.4, and the color can be matched to pink; the trees in the forest are dense, the ground is covered with fallen leaves and weeds, and the visibility is limited. The trafficability of the vehicle is recorded as 0.2, and the color can be matched to dark green; the water system includes rivers, lakes and wetlands, and vehicles cannot pass through, which is recorded as 0, and the color can be matched to blue.

[0070] As Figure 3 shown, the basic attributes of the raster overlay map include: raster abscissa, raster ordinate, elevation information, raster slope and surface attributes; it can be expressed as G xy =[x, y, h, s, b]; where x represents the raster abscissa, y represents the raster ordinate, h represents the elevation information; s represents the raster slope; b represents the surface attribute. Information such as obstacles, risk areas and passable areas is described through these basic attributes.

[0071] The raster slope of each raster is calculated based on the basic raster map. It should be noted that the slope is a simplified form of the actual surface curvature. To simplify the problem, the calculation method of the raster slope in the present invention includes: regarding a small piece of the surface curvature corresponding to each raster as a plane, and the dihedral angle value between the plane and the horizontal plane is the raster slope.

[0072] Furthermore, the calculation formula for the raster slope s is:

[0073]

[0074] Where represents the slope change rate of the current raster in the x direction of the raster abscissa; represents the slope change rate of the current raster in the y direction of the raster ordinate.

[0075] As Figure 4 shown, the eight rasters adjacent to raster 5 are 1, 2, 3, 4, 6, 7, 8, 9 respectively. The slope change rate of raster 5 in the x direction of the raster abscissa and the slope change rate in the y direction of the raster ordinate are respectively:

[0076]

[0077] Among them, h(1), h(2), h(3), h(4), h(6), h(7), h(8), and h(9) respectively represent the elevations of grids 1, 2, 3, 4, 6, 7, 8, and 9; size represents the length of the grid.

[0078] Facing the grid slope, a smoothness index can be adopted to measure the bumpiness of each grid slope during the path driving of the unmanned vehicle.

[0079] It should be noted that the calculation method of the smoothness index e is as follows:

[0080]

[0081] s l = |s i - s k |;

[0082] Among them, s l represents the slope difference between two adjacent grids i and k in the driving path; s i , s k respectively represent the slope differences between the adjacent grid i and grid k in the driving path.

[0083] In the entire driving path of the unmanned vehicle, the lower the smoothness index, the better the smoothness of the path.

[0084] For the entire grid overlay map including an m×n grid area, it can be represented as a three-dimensional array G mn = G xy = [x, y, h, s, b].

[0085] Step 103: Improve the A* algorithm using the neighborhood search method to obtain the improved A* algorithm.

[0086] It should be noted that the A* algorithm is a heuristic search algorithm widely used in path planning and graph search problems; it combines the global optimality of the Dijkstra algorithm and the efficiency of the greedy best-first search, and guides the search process by using a heuristic evaluation function, so as to effectively find the shortest path from the starting node to the target node on the premise of meeting optimality and completeness.

[0087] The basic idea of the A* algorithm is to maintain a priority queue and dynamically evaluate the "cost" of each node; the core of the algorithm lies in defining a comprehensive evaluation function f(n) to evaluate the total estimated cost from the starting grid to the current grid n. The function of the A* algorithm mainly consists of two parts:

[0088] f(n) = g(n) + h(n);

[0089] Among them, g(n) represents the actual cost from the starting grid p to the current grid n; h(n) represents the estimated cost from the current grid n to the destination grid q.

[0090] As an embodiment, improving the A* algorithm using the neighborhood search method includes: expanding the search neighborhood to multiple layers of grids centered on the current grid and extending outward, so that the path search can not only search the first-layer grids adjacent to the current grid, but also search the second-layer grids adjacent to the first-layer grids. This is a dynamic adjacency mechanism based on a more extensive direction. By implementing a more extensive search direction, the flexibility and reliability of path planning are significantly improved.

[0091] Specifically, as Figure 5 shown, the neighborhood search method can search the first-layer grids adjacent to the current grid and the second-layer grids adjacent to the first-layer grids, including searching 8 directions in the first layer and 16 directions in the second layer. That is to say, the neighborhood search method expands the search domain to 24 directions; making the path search not only able to consider the directly adjacent grids around the current grid, but also able to cover farther neighborhood nodes. By introducing additional search directions, the mutations in the path can be effectively reduced, and the continuity and smoothness of the overall path can be improved.

[0092] Since when searching in the wild environment, there is a large amount of environmental information to be processed, the search direction cannot be simply pruned based on the relative position relationship between the destination grid and the current grid. The dynamic adjacency mechanism dynamically adjusts the search neighborhood by real-time evaluating the changes in the environment, such as obstacles, risk areas, and passability.

[0093] Therefore, in different environmental states, in order to enable the improved A* algorithm to flexibly select the most suitable adjacent grid for path search according to the characteristics of the current grid and the status of surrounding resources. As an embodiment, when performing neighborhood search, the conditions for generating neighborhood grids need to satisfy the first condition and the second condition simultaneously; among them, the first condition is: the neighborhood grid is not occupied by obstacles or is a passable area; the second condition is: the neighborhood grid is within the range of the grid overlay map.

[0094] In other words, when generating neighborhood grids, there are mainly two factors affecting the passability of grids. One is to check whether the generated neighborhood grid is occupied by obstacles or is an inoperable area. If so, the neighborhood grid is unavailable; the other is to ensure that the generated neighborhood grid is within the boundary of the grid map to avoid accessing invalid grids.

[0095] Step 104: Determine the starting grid and the destination grid in the grid overlay map, and use the improved A* algorithm to perform path planning on the grid overlay map to obtain the optimal path from the starting grid to the destination grid.

[0096] Traditional heuristic functions often only consider geometric distance and timeliness, but do not fully consider the surface properties, stability and potential risks of the path.

[0097] In order to provide different path planning solutions for different complex field environments or tasks and meet the requirements of real field environments and different unmanned vehicle tasks, the present invention constructs one or more priority heuristic functions based on timeliness, safety, and stability for different complex field environments, and uses the improved A* algorithm to perform path planning on the grid overlay map. The present invention not only considers the influence of surface features and risk areas, but also designs different heuristic functions for different scenario requirements in order to provide more options, rather than simply providing a path planning strategy.

[0098] In the path planning of unmanned vehicles, timeliness refers to the minimization of the time cost required for the vehicle to complete the path planning task. It directly reflects the operating efficiency of the path and the speed of task completion. Especially in complex outdoor environments, timeliness puts higher requirements on the path planning of unmanned vehicles.

[0099] Specifically, when the task is of high urgency (such as disaster rescue, emergency material distribution and other scenarios), the unmanned vehicle needs to choose a path with the shortest time and the highest mission completion efficiency; at the same time, due to the particularity of the field environment, different terrains (such as mud, sand, rocky roads or flat grass) have a significant impact on the vehicle's driving speed, which makes it difficult to meet actual needs if the planning strategy is based solely on the shortest path length.

[0100] As an example, the first priority heuristic function f t (n) is:

[0101] f y (n) = g t (n)+h t (n);

[0102]

[0103] Where time(i) represents the time required to travel in the grid; d n Indicates the distance between the current grid n and the target grid; v max Indicates the maximum speed of the driverless car.

[0104] Safety is a crucial indicator in the path planning of unmanned vehicles, which is directly related to the feasibility of the path and the risk control during the task execution. Especially in complex outdoor environments, unmanned vehicles need to face potential threats with strong dynamics and high uncertainty, such as irregular terrain undulations and hidden obstacles (such as rocks or tree roots in the grass).

[0105] Therefore, path planning not only needs to focus on task efficiency, but also must ensure that the unmanned vehicle avoids potential dangers throughout the driving process, maximizing task safety.

[0106] As an example, the second-priority heuristic function f r (n) is:

[0107] f r (n) = g r (n) + h r (n);

[0108]

[0109] h r (n) = d n ;

[0110] where risk(i) represents the risk time of driving in the grid; d n represents the distance between the current grid n and the destination grid.

[0111] In the path planning of unmanned vehicles, the core of the smoothness index is to evaluate the smoothness of the path during vehicle driving and its stability on complex terrains. This index is directly related to the driving comfort, stability, and equipment safety of unmanned vehicles. Especially in the wild and complex environments, an uneven path may lead to errors in the navigation system, accelerated wear of vehicle hardware, and even significantly affect the success rate of task execution.

[0112] In view of the path planning requirements of unmanned vehicles in the wild environment, based on the elevation information in the grid overlay map, the present invention further determines the slope difference between adjacent grids on the path, and at the same time proposes a smoothness measurement method centered on the sum of the slope differences of all adjacent grids on the path, realizing the quantitative evaluation and optimization of path smoothness.

[0113] As an example, the third-priority heuristic function f s (n) is:

[0114] f s (n) = g s (n) + h s (n);

[0115]

[0116] h s (n) = d n ·avg;

[0117]

[0118] Among them, s l represents the slope difference between two adjacent grids in the driving path; avg represents the average slope difference of the driving distance of the driverless vehicle.

[0119] It should be noted that the priority heuristic functions of timeliness, safety, and smoothness can be measured separately or together, and the result is obtained after weighted averaging.

[0120] As an example, the measurement index of the optimal path is:

[0121] f(n) = ω t f t (n) + ω r f r (n) + ω s f s (n);

[0122] Among them, ω t , ω r and ω s represent the weight values of timeliness, safety, and smoothness in sequence. The setting of the weight values depends on the task requirements of the driverless vehicle in the wild environment, whether it pays more attention to timeliness, safety, or smoothness.

[0123] According to the Euclidean distance calculation method, combined with the actual complex wild environment situation, the present invention considers the influencing factors of elevation information, and the calculation method of the diameter distance from the starting grid to the current grid can be obtained as:

[0124]

[0125] Among them, x p , y p , h p represent the grid abscissa, grid ordinate, and elevation information of the starting grid p; x n , y n , h n represent the grid abscissa, grid ordinate, and elevation information of the current grid n.

[0126] In the second aspect, the present invention provides a driverless vehicle path planning system applicable to a complex environment, including a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of the method described in any one of the above. Details are not described herein again.

[0127] It should be noted that for the foregoing embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to this application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0128] In the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0129] In the several embodiments provided by this application, it should be understood that the disclosed method or system can be implemented in other ways. For example, the above-described embodiments are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0130] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0131] In addition, in each embodiment of this application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0132] If the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of this application.

[0133] Those of ordinary skill in the art can understand that all or part of the circuits in the above embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable memory. The memory may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, an optical disk, etc.

[0134] The above are only exemplary embodiments of the present disclosure, and the scope of the present disclosure cannot be limited thereby. That is, any equivalent changes and modifications made in accordance with the teachings of the present disclosure still fall within the scope covered by the present disclosure. After considering the specification and practicing the present disclosure herein, those skilled in the art will readily conceive of other embodiments of the present disclosure. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include known common general knowledge or conventional technical means in the technical field not described in the present disclosure. The specification and embodiments are only regarded as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

[0135] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. As long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0136] Those skilled in the art can easily understand that the above are only preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A path planning method for an unmanned vehicle in a complex environment, characterized in that: The method comprises: Obtain basic grid maps of complex outdoor environments; Calculating the grid slope of each grid according to the basic grid map; setting different surface attributes according to different surface information in the basic grid map and the passability of the unmanned vehicle on different surface information; superimposing the grid slope and the surface attributes on the elevation information layer of the corresponding grid in the basic grid map to obtain a grid superposition map; The A* algorithm is improved by using the neighborhood search method to obtain the improved A* algorithm; A starting grid and a destination grid are determined in the grid overlay map, and a path planning is performed on the grid overlay map using an improved A* algorithm to obtain an optimal path from the starting grid to the destination grid.

2. The unmanned vehicle path planning method applicable to complex environments according to claim 1, characterized in that: Setting different surface properties includes: Different color codes and / or texture codes are matched for each grid in the basic grid map according to the passability corresponding to different surface information.

3. The unmanned vehicle path planning method applicable to complex environments according to claim 1, characterized in that: The basic properties of the grid overlay map include: grid abscissa, grid ordinate, elevation information, grid slope and surface properties.

4. The method for unmanned vehicle path planning applicable to complex environments according to claim 1, characterized in that: The method for calculating the grid slope includes: considering a small surface of the ground corresponding to each grid as a plane, and the dihedral angle value between the plane and the horizontal plane is the grid slope.

5. The unmanned vehicle path planning method applicable to complex environments according to claim 4, characterized in that: The calculation formula of the grid slope s is: in, Indicates the slope change rate of the current grid in the x-direction of the grid horizontal coordinate; Indicates the slope change rate of the current grid in the y direction of the grid vertical coordinate.

6. The method for unmanned vehicle path planning applicable to complex environments according to claim 1, characterized in that: Improvements to the A* algorithm using the neighborhood search method include: The search neighborhood is expanded to multiple layers of grids extending outward from the current grid as the center, so that the path search can search not only the first layer of grids adjacent to the current grid, but also the second layer of grids adjacent to the first layer of grids.

7. The method for unmanned vehicle path planning applicable to complex environments according to claim 6, characterized in that: When performing a neighborhood search, the conditions for generating a neighborhood grid must satisfy both the first and second conditions; The first condition is that the neighborhood grid is not occupied by obstacles or is a passable area; The second condition is that the neighborhood grid is within the range of the grid overlay map.

8. The unmanned vehicle path planning method applicable to complex environments according to claim 1, characterized in that: Aiming at different complex outdoor environments, one or more priority heuristic functions based on timeliness, safety, and stability are constructed, and the improved A* algorithm is used to perform path planning on the grid overlay map.

9. The method for unmanned vehicle path planning applicable to complex environments according to claim 8, characterized in that: The first priority heuristic function f of the timeliness t (n) is: f t (n)=g t (n)+h t (n); Where time(i) represents the time required to travel in the grid; d n Indicates the distance between the current grid n and the target grid; v max Indicates the maximum speed of the unmanned vehicle; The second priority heuristic function f of the security r (n) is: f r (n)=g r (n)+h r (n); h r (n)=d n ; Among them, risk(i) represents the risk time of driving in the grid; d n Indicates the distance between the current grid n and the target grid; The third priority heuristic function f for stationarity s (n) is: f s (n)=g s (n)+h s (n); h s (n)=d n ·avg; Among them, s l It represents the slope difference between two adjacent grids in the driving path; avg represents the average slope difference of the unmanned vehicle's driving distance.

10. A path planning system for an unmanned vehicle in a complex environment, comprising a memory, a processor, and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 9.

Images