Research On Path Planning Algorithm Of UAV

In recent years,with the improvement of the intelligence of UAVs,UAVs have been widely used in military,rescue,detection and other related fields.UAV-related technologies have also become the subject of extensive research by many scholars at home and abroad,among which how to efficiently and safely plan the flight path of UAVs has become a current research hotspot,so it is extremely important to develop efficient and safe path planning algorithms for the autonomous flight of UAVs.This paper conducts research on the UAV path planning algorithm.First,the Bi-directional Rapidly Exploring Random Tree(Bi-RRT)algorithm has problems such as blind search,long and tortuous paths in global path planning.Secondly,aiming at the local path planning problem,an improved dynamic window method is adopted,and the global path planning algorithm is fused with the local path planning algorithm.Finally,the fusion improved algorithm is deployed on the flight control platform based on Pixhawk for real aircraft testing.The main work and innovations of this paper are as follows:1.Aiming at the global path planning problem of UAVs in a static environment,an improved Bi-RRT fusion Dijkstra algorithm global path planning algorithm is designed.First,the constrained probability sampling and goal-oriented growth strategy are used to improve the search efficiency of the Bi-RRT algorithm.Then,the redundant path points in the path are removed by combining the Dijkstra algorithm.Finally,the optimized path is smoothed to obtain an optimal feasible path.In the two-dimensional simulation environment,several experiments were carried out on the fusion improved algorithm.The experiments showed that the search efficiency of the fusion improved Bi-RRT algorithm was improved,the path search time was reduced by 32.97%,and the path length was shortened by 7.97%.In addition,the fusion improved Bi-RRT algorithm is extended to the three-dimensional environment space.Experiments show that the search efficiency of the improved Bi-RRT algorithm is improved,the path search time is reduced by 21.46%,and the path length is shortened by 14.75%.2.For the local path planning problem of UAVs in a dynamic environment,firstly,by introducing the influence of the distribution rate of environmental obstacles,and then dynamically adjust the coefficient value of the azimuth evaluation function of the dynamic window approach,so as to avoid the algorithm from easily falling into the local optimal solution;secondly The starting point pose optimization strategy is used to improve the problem of local path redundancy in dynamic path planning;finally,the improved dynamic window approach is combined with the improved global path planning algorithm.This enables the algorithm to ensure the optimal global path and at the same time perform local dynamic path planning in a dynamic environment.3.Based on the above research on the path planning algorithm,a set of software simulation and hardware UAV autonomous path planning system has been built independently.The software simulation system is a real machine simulation platform based on the ROS operating system,and the hardware platform is based on Pixhawk flight control.Finally,the autonomous path planning test is carried out on the software simulation platform and the outdoor environment respectively.After testing,the UAV autonomous path planning system built in this paper can effectively realize autonomous path planning and obstacle avoidance flight.