UAV Task Planning Based On Intelligent Optimal Algorithms

With the progress of science and technology,especially the rapid development of high and new technology,major changes have taken place in the military field.Military UAVs integrate aerospace technology,intelligent information processing technology,communication technology,precision locking technology and strike technology.And military UAVs began to emerge in the modern battlefield.In recent years,research on task planning of multi-UAV has become the hot point of research.The key problem to be solved in the multi-UAV task planning is how to allocate tasks reasonably for multiple UAVs in complex and changeable battlefield environment and plan a safe flight path for them from the starting point to the target point,which can avoid radar,threat area and obstacles.The main contents of this paper are as follows:(1)This paper proposes a multi-UAV task allocation based on improved algorithm of multi-objective particle swarm optimization.This algorithm uses two-stage Monte Carlo sampling process to improve the particle swarm optimization algorithm on the shortcomings of easy falling into local optimum solution and slow search speed in the middle and late stages,so that the algorithm can search for more non-dominated solutions.(2)This paper proposes a task reassignment method for UAV based on K-medoids clustering algorithm.K-medoids clustering algorithm is used to cluster UAVs and mission points in battlefield into clusters with the same number of UAVs.Each cluster is solved as a single traveling salesman problem.In this way,the dimensions of original problem and the running time of the algorithm can be greatly decreased.(3)This paper proposed an UAV route planning algorithm based on improved probabilistic roadmap algorithm and ant colony algorithm.The improved probabilistic roadmap algorithm pretreats the flight environment and converts it into a connectionless graph.In improved probabilistic roadmap algorithm,number of points picked near obstacles is increased to enhance the path search in narrow channels.The improved probabilistic roadmap algorithm is more suitable for complex terrain.The shortest path in the connectionless map is solved by ant colony algorithm.The experimental results show that the task assignment method proposed in this paper can effectively pre-allocate tasks for multiple UAVs,and can greatly reduce the time for UAVs to reassignment tasks when encountering unexpected situations.The UAV route planning algorithm can be adapted to more complex environments and improve the adaptability of route planning.