An improved a* algorithm-based adaptive path planning method for unmanned aerial vehicles

Abstract

The application discloses a kind of unmanned plane self-adapting path planning method based on improved A* algorithm, first, a new heuristic cost considering the obstacles in environment is designed for hybrid A* algorithm, and considering that unmanned plane flies in a chaotic degree uniform map, then an adaptive strategy is designed to find the optimal heuristic cost for the current specific environment.The scheme can enable unmanned plane to find the optimal inflation coefficient corresponding to the map environment after executing hybrid A* algorithm for more than ten times, and after obtaining the value, the unmanned plane normally executes hybrid A* algorithm with the optimal inflation coefficient to quickly find the optimal path.

Description

TECHNICAL FIELD

[0001] The present application relates to a path planning method, in particular to an improved A* algorithm-based adaptive path planning method for unmanned aerial vehicles. BACKGROUND

[0002] With the development of technology, unmanned aerial vehicles have shown great potential in many fields. Among the many related technologies that support the use of unmanned aerial vehicles, path planning technology is an important part. The efficiency of the algorithm has a great impact on the response speed and safe flight of the unmanned aerial vehicle. Therefore, it is of great significance to design a fast and efficient adaptive path planning method for the flight environment of unmanned aerial vehicles.

[0003] Traditional path planning methods based on hybrid A* algorithm will try to speed up the path search as much as possible by designing heuristic functions. In the prior art, a hybrid A* method is proposed to plan the path, and the Euclidean distance between the current point and the target point is used as the heuristic cost. This heuristic cost does not take into account the existence of obstacles, resulting in low search efficiency. The second heuristic cost construction method uses the Dijsktra algorithm on the global map to calculate the shortest safe distance from the target point to each point on the map. However, this method needs to use the Dijsktra algorithm constantly as the map updates, which has a large computational burden and cannot achieve real-time performance. The existing method also calculates a two-point boundary value problem from the current state point to the target point, models it as an optimization problem and uses the Pontryagin minimum principle to obtain a closed-form solution. The cost corresponding to the optimal solution is used as the heuristic cost. Although this method takes into account the dynamic characteristics of the unmanned aerial vehicle and is closer to the real cost than the Euclidean distance, it still does not take into account the existence of obstacles, and the heuristic cost obtained is still less than the real cost, so the path planning speed is still not fast enough. SUMMARY

[0004] In view of the above deficiencies in the prior art, the present application provides an improved A* algorithm-based adaptive path planning method for unmanned aerial vehicles.

[0005] In order to achieve the above-mentioned application purposes, the technical solution adopted by the present application is as follows:

[0006] An improved A* algorithm-based adaptive path planning method for unmanned aerial vehicles, comprising the following steps:

[0007] S1. Flying the unmanned aerial vehicle in a map with uniform chaos, and discretizing the map after sensing it;

[0008] S2. Constructing an adaptive strategy and setting an optimal heuristic cost for the adaptive strategy according to the obstacles in the current environment;

[0009] S3, execute the adaptive strategy, calculate the optimal inflation coefficient corresponding to the current map environment;

[0010] S4, use the obtained optimal inflation coefficient to execute the hybrid A* algorithm to find the optimal path.

[0011] Further, the S2 specifically comprises the following steps:

[0012] S21, considering the existence of obstacles, solve the optimal two-point boundary value from the current state point to the target point;

[0013] S22, calculate the proportion of the part of the optimal trajectory colliding with the obstacle according to the obstacle grid map;

[0014] S23, correct the heuristic cost according to the proportion calculated in step S22 to obtain the optimal heuristic cost. Further, the optimal heuristic cost in S23 is represented as:

[0015] J ′ =(1-)J+I·αJ

[0016] Where, j ′ is the optimal heuristic cost, a is the proportion of the trajectory collision part to the entire optimal trajectory, I is the inflation coefficient, and J is the current optimal heuristic cost.

[0017] Further, the S3 specifically comprises the following steps:

[0018] S31, execute the first stage of the adaptive strategy to find the optimal interval of the inflation coefficient;

[0019] S32, execute the second stage of the adaptive strategy to narrow the interval range on the basis of step S31 to approximate the optimal inflation coefficient.

[0020] Further, the S31 specifically comprises the following steps:

[0021] S311, initialize the inflation coefficient and the expansion point of the A* algorithm;

[0022] S312, judge whether the expansion point number corresponding to the right boundary of the inflation coefficient is greater than the expansion point number corresponding to the intermediate value, if greater than the intermediate value expansion point number, then enter the next stage; if less than the intermediate value expansion point number, then update the values of the left boundary, the intermediate value and the corresponding expansion point number respectively;

[0023] S313, update the value of the right boundary using the set step length and execute the A* algorithm repeatedly, and take the finally obtained expansion point number as the expansion point number corresponding to the right boundary.

[0024] Further, the S32 specifically comprises the following steps:

[0025] S321. Initialize the step size;

[0026] S322. Determine if the interval length is greater than the threshold. If it is less than the threshold, end the loop and the middle node is the optimal expansion coefficient. If it is greater than the threshold, shrink the interval and divide the left and right boundaries into two parts. Move the left and right boundaries one step towards the middle and execute the hybrid A* algorithm to obtain the number of expansion points under the current expansion coefficient.

[0027] S323. Compare the newly obtained number of expansion points with the number of expansion points of the original left and right boundaries of the interval. Update the left and right boundaries of the interval according to different size relationships to shrink the interval. After the update is completed, halve the step size and repeat step S322.

[0028] The present invention has the following beneficial effects:

[0029] 1) The impact of obstacles in the environment on pathfinding methods was fully explored, and a heuristic function that is closer to the actual flight cost was designed, which greatly improved the speed of path search.

[0030] 2) Fully recognizing the consistency of obstacle characteristics in the mission environment of the UAV, an adaptive strategy was designed to find the optimal expansion coefficient corresponding to the specific environment, so that the heuristic cost of the A* algorithm approaches its optimal upper limit, greatly reducing the expansion points of the A* algorithm. Attached Figure Description

[0031] Figure 1 This is an overall flowchart of the path planning method described in this invention.

[0032] Figure 2 This is a flowchart of step 2 of the present invention.

[0033] Figure 3 This is a schematic diagram of the calculation and correction of the trajectory prediction in step 2 of the present invention, which involves the collision of the original trajectory.

[0034] Figure 4 This is a flowchart illustrating the process of finding the optimal interval in step 3 of this invention.

[0035] Figure 5 This is a flowchart illustrating the process of finding the optimal expansion coefficient in step 3 of this invention. Detailed Implementation

[0036] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0037] An adaptive path planning method for unmanned aerial vehicles (UAVs) based on an improved A* algorithm, such as... Figure 1 As shown, it includes the following steps:

[0038] S1. Use a drone to fly over a map with uniform disorder, perceive the map, and then discretize it.

[0039] Drones typically operate within relatively small, localized environments, such as forests, offices, logistics warehouses, and urban buildings. In these environments, the geometric features and density of obstacles are relatively uniform and do not change significantly; therefore, such environments can be considered uniformly chaotic. We use LiDAR or depth cameras to acquire obstacle information about the drone's environment, and then create an obstacle grid map based on the point cloud data, saving it to a discrete grid map module.

[0040] S2. Construct an adaptive strategy and set an optimal heuristic cost for the adaptive strategy based on the obstacles in the current environment;

[0041] Considering the presence of obstacles, a new heuristic cost function is designed for the hybrid A* algorithm. The specific process is shown in the attached figure. Figure 2 As shown. First, we need to solve an optimal two-point boundary value problem (OBVP) from the current state point to the target point: model a convex optimization problem with minimum path cost, and use the Pontryagin minimization principle to solve it, to obtain the optimal trajectory and the corresponding optimal cost J.

[0042] Then, based on the obstacle grid map, the proportion of the optimal trajectory that collides with obstacles is calculated. The trajectories that collide with obstacles are shown in the attached figure. Figure 3 As shown by the thick solid line, the specific calculation method for the collision trajectory involves discrete sampling over time, calculating the trajectory position at each sampling moment, and then querying the grid map module to determine whether the grid where the trajectory position is located is an obstacle grid. Let α be the proportion of the trajectory collision part to the entire optimal trajectory (0 <= α <= 1).

[0043] The collision trajectory will not be the actual flight trajectory of the drone; the actual flight trajectory should be similar to the one shown above. Figure 3Either trajectory ① or trajectory ②, regardless of which trajectory it is, its minimum cost must be greater than the current optimal cost J, and it is related to α. In this invention, we modify the optimal cost according to the collision ratio as follows:

[0044] J ′ =(1-)J+I·αJ

[0045] Here, I (I>=1) is the inflation factor, which represents how many times the original cost J should be inflated according to the collision ratio α. This method makes the corrected heuristic cost closer to the actual path cost, bringing the heuristic cost h closer to its optimal upper bound h*, significantly improving the convergence rate of the A* algorithm. Since the inflation factor I is related to the geometric features of the obstacles, each specific map should correspond to a different, local optimal inflation factor. An adaptive strategy is then used to find the optimal inflation factor for each specific map.

[0046] S3. Execute the adaptive strategy and calculate the optimal expansion coefficient corresponding to the current map environment;

[0047] This embodiment specifically includes the following steps:

[0048] S3 specifically includes the following steps:

[0049] S31. The first stage of implementing the adaptive strategy is to find the optimal range of the inflation coefficient.

[0050] like Figure 4 As shown, the left and right boundaries and the middle position of the expansion coefficient interval are first initialized, specifically the expansion coefficient and the number of expansion points of the A* algorithm. Then, a loop is entered. The loop entry condition is to determine whether the number of expansion points corresponding to the right boundary of the expansion coefficient is greater than the number of expansion points corresponding to the middle value. If it is greater than the number of expansion points corresponding to the middle value, the next stage is entered; if it is less than the number of expansion points corresponding to the middle value, the values ​​of the left boundary and the middle value and the corresponding number of expansion points are updated respectively. Then, the value of the right boundary is updated using the step size, the hybrid A* algorithm is executed, and the number of expansion points finally obtained by the algorithm is used as the number of expansion points corresponding to the right boundary value. Then, the loop condition is returned.

[0051] S32. The second stage of implementing the adaptive strategy involves narrowing the interval range based on step S31 to approximate the optimal expansion coefficient.

[0052] In the second stage of the adaptive strategy, based on the interval obtained in step S31, the range is further narrowed to approximate the optimal inflation coefficient. The specific process for this step is attached. Figure 5As shown, the step size is first initialized, typically to 10. Then, the first loop condition is entered: check if the interval length is greater than a threshold. If it is less than the threshold, the loop ends, and the middle node is the optimal expansion coefficient. If it is greater than the threshold, the interval is shrunk by moving the left and right boundaries one step towards the middle, and the hybrid A* algorithm is executed to obtain the number of expansion points under the current expansion coefficient. Then, the newly obtained number of expansion points is compared with the original number of expansion points on the left and right boundaries of the interval. The left and right boundaries of the interval are updated according to different size relationships, shrinking the interval. After the update is complete, the step size is halved, and the next loop begins.

[0053] S4. Use the obtained optimal expansion coefficient to execute the hybrid A* algorithm to find the optimal path.

[0054] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0056] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0057] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

[0058] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. An improved A* algorithm-based unmanned aerial vehicle adaptive path planning method, characterized in that, The method comprises the following steps: S1, flying on a uniform chaos map by using a UAV, and discretizing the map after sensing the map; S2, constructing an adaptive strategy and setting an optimal heuristic cost for the adaptive strategy according to obstacles in a current environment; S3, executing the adaptive strategy, and calculating an optimal inflation coefficient corresponding to a current map environment, comprising the following steps: S31, executing a first stage of the adaptive strategy, and searching for an optimal interval of the inflation coefficient, comprising the following steps: S311, initializing the inflation coefficient and an expansion point of the A* algorithm; S312, judging whether the expansion point number corresponding to the right boundary of the inflation coefficient is greater than the expansion point number corresponding to the intermediate value, if greater than the intermediate value expansion point number, entering a next stage; if less than the intermediate value expansion point number, updating the values of the left boundary and the intermediate value and the corresponding expansion point numbers respectively; S313, updating the value of the right boundary by using a set step length repeatedly according to step S312, and executing the A* algorithm, and taking the final expansion point number as the expansion point number corresponding to the right boundary; S32, executing a second stage of the adaptive strategy, and narrowing the interval range on the basis of step S31 to approximate the optimal inflation coefficient, S321, initializing the step length; S322, judging whether the interval length is greater than a threshold value, if less than the threshold value, ending the loop, and the intermediate node being the optimal inflation coefficient; if greater than the threshold value, narrowing the interval, and moving the left and right boundaries to the middle by one step length, and executing the hybrid A* algorithm respectively to obtain the expansion point number under the current inflation coefficient; S323, comparing the newly obtained expansion point number with the expansion point numbers of the left and right boundaries of the original interval according to different size relationships, updating the left and right boundaries of the interval respectively, narrowing the interval, and halving the step length after updating and repeating step S322; S4, executing the hybrid A* algorithm by using the obtained optimal inflation coefficient to find an optimal path.

2. The improved A* algorithm-based UAV adaptive path planning method according to claim 1, characterized in that, The S2 specifically comprises the following steps: S21, considering the existence of obstacles, and solving optimal two-point boundary values from a current state point to a target point; S22, calculating a proportion of a part where an optimal trajectory collides with obstacles according to the obstacle grid map; S23, correcting the heuristic cost according to the proportion calculated in step S22 to obtain an optimal heuristic cost.

3. The improved A* algorithm-based UAV adaptive path planning method according to claim 2, characterized in that, The optimal heuristic cost in step S23 is expressed as: wherein, is the optimal heuristic cost, is the ratio of the trajectory collision portion to the entire optimal trajectory, is the inflation coefficient, is the current heuristic cost.

Images