Research On Path Planning Algorithm For UAV Under Multi Constraints

The technology of the route planning is always the research hotspot in the UAVs world.In this thesis, after analyzed the research status at the inland or at the foreign countries. In this thesis, the constraints are divided into three categories, which are the external environment constraints, the UAV’s performance constraints and the task constraints. The external environment constraints include the natural terrain obstacles, the weather conditions, the radar position and the missile position and so on. The performance constraints of UAV include the most air speed of the UAVs, the minimum turning radius, and the farthest voyage of the UAVs.Moreover the task condition is based on the task which UAVs were processing.In this thesis, I use the grid method to build the environment model for UAVs. And use the A* algorithm as the search algorithm of the UAV route planning. Because of the A*algorithm have some weaknesses, we improved the algorithm, and compare with the traditional A* algorithm and the bidirectional A * algorithm. Meanwhile, in this thesis, we introduce a new synchronous bidirectional A* algorithm to solve the problem that the forward search and backward search are always waiting for each other. In the 3D environment, we compared the performance between the A* algorithm and particle swarm optimization algorithm.At last, with the simulation experiments by using the traditional A* algorithm, the bidirectional A* algorithm and the synchronous bidirectional A* algorithm. The performance of three algorithms are compared in 2D environment, and the experiment results shows that the synchronous bidirectional A* algorithm has a great improvement in the search speed. In the 3D environment, we make simulation experiments by using the bidirectional A* algorithm and particle swarm optimization algorithm, and compare the performance of the two algorithms. The experiment results shows that the performance of PSO in 3D environment is better than the A* algorithm,we analyzed the reason why the A* algorithm will become slow in the 3D environment, and point out that the stochastic search algorithms are more efficiencyin the 3D environment.