基于差分进化分布估计算法的无人机航迹规划  

The Track Planning of Unmanned Aerial Vehicle based on the DE-EDA

作　　者：黄艳虎 李天松[1,2] 卢亚军 阳荣凯[1] 

机构地区：[1]桂林电子科技大学信息与通信学院,广西桂林541004 [2]广西精密导航技术与应用重点实验室,广西桂林541004

出　　处：《计算机仿真》2018年第1期23-26,共4页Computer Simulation

基　　金：国家自然科学基金(61263028);广西高校科学技术研究项目(KY2015ZD042);桂林电子科技大学研究生教育创新计划资助项目(YJCXS201520);国家大学生创新创业训练计划资助项目(201410595010)

摘　　要：无人机航迹规划问题的研究中,为了达到撞机概率最小、航迹搜索速度最快、航迹长度最短,提出了一种基于差分进化-分布估计(DE-EDA)算法相结合的无人机多目标最优航迹规划方法。首先,建立简单的三维模拟飞行环境;其次,通过飞行约束条件构造航迹长度、威胁指数多目标优化模型,把动态航迹规划问题转化为动态优化问题;最后,采用DE-EDA组合算法平衡航迹长度与威胁指数不同资源分配的情况,寻找一组适中解,确定一条最优航迹。实验表明:飞行航迹由于外界环境变化而变为非优或者不可行时,上述方法能够快速调整寻优策略,并对其它次优路径进行寻优,即使在障碍物复杂的环境中,对于多目标点问题,上述算法仍能规划出较好的目标遍历路径,且用时较少。ABSTRACT：This paper makes a research for Unmanned Aerial Vehicle track planning problems, which proposes an Unmanned Aerial Vehicle multi-objective track planning method based on the combination of DE-EDA under the complex environment to realize minimum impact probability, fastest route research and minimum cost for track. At first, we established a simple 3D flight simulation environment. Second, we made up track length by the constraint conditions of flight to threat cost for multi-objective optimization and turn the dynamic track planning problem into a dynamic optimization problem. Finally, the combination of DE-EDA was used to balance the allocation for different resources which are between track length and threat cost to look for a group of moderate solution. Experiment has show that due to the external environment, the flight track becomes optimal or not applicable, but this method can raoidlv adjust the ootimization strategy and make an ontimization for other subontimal oath to ~enerate a oerfect track.

关 键 词：无人机 航迹规划 多目标优化 差分进化-分布估计 

分 类 号：V271.4[航空宇航科学与技术—飞行器设计]