基于MSMPCE-GA的多无人机协同侦察路径规划研究OA

Aimed at the problems of strong dynamic uncertainty and high-dimensional search spaces in the process of multiple unmanned aerial vehicle(multi-UAV)cooperatively reconnoitering maneuvering targets,a multi-UAV cooperative reconnaissance path planning method based on a multi-scale multi-population co-evolu-tion genetic algorithm(MSMPCE-GA)is proposed.First,based on the mobile target distribution prior,the discov-ery probability of multiple reconnaissance attempts by multi-UAVs is derived,then a path planning model for multi-UAV cooperative reconnaissance is constructed with the goal of maximizing the probability and simultane-ously considering flight range and time constraints,and finally,a multi-scale search space is employed for model-ing,and the MSMPCE-GA algorithm with multi-population co-evolution mechanism is for solution.The simula-tion results show that,compared with the standard genetic algorithm(SGA),the method proposed in the paper has the average target discovery probability increased from 78.97%to 91.93%,capable of improving the search effi-ciency of multi-UAV cooperative reconnaissance,thus providing a feasible solution for cooperative reconnais-sance in complex dynamic environments.