多无人机协同任务规划问题

研究无人机任务规划问题,从无人机侦查和轰炸两方面入手.首先,运用迭代算法求解出从基地到雷达区域边际上任一目标出入口的最短路径.在此基础上,以无人机在雷达范围内滞留时间最短,以及被探测次数最少为目标,建立多目标最优化模型.通过改进交叉算子的遗传算法找出最优侦查路径.对于轰炸任务,以无人机在雷达范围内滞留时间最短,以及轰炸总时间最短为目标,建立多目标最优化模型.采用改变惯性权重的自适应粒子群算法找出最佳轰炸路线.由于计算时间较长,本文对68个目标进行聚类分析,提出针对轰炸任务的快速算法,相较原轰炸方案,其计算效率提高80%以上.

Mission Planning of Multiple Unmanned Aerial Vehicles

This paper researches mission planning of unmanned aerial vehicles (UAV),analyzing in two parts,investigation and attack.Firstly,communicate the shortest path between bases and boundaries of radar cover region by iteration.Based on this,build a multi-objective optimization model with objects,the least time staying in cover region and the least times being probed for.The most optimal path can be found by general algorithm with improved crossovers.In regard of attack missions,a multi-objective optimization model is built with two objects,the least time staying in cover region and the shortest attacking time.The most optimal path is found by particle swarm optimization with changing weight.Considering that calculation costs much time,68 targets are divided into different classifications.An improved fast algorithm aiming at attacking mission was put forwards,which is 80％ more effective than the original.