改进遗传算法的无人机路径规划

UAV path planning based on improved genetic algorithm

为了进一步研究无人机(UAV)在自主飞行任务中的飞行安全问题,在保证飞行路程尽可能短的基础上,提出1种基于改进遗传算法的路径规划方法:给出使无人机飞行轨迹远离障碍物的思想,以减少因环境感知与飞行控制误差发生碰撞的机会:在种群初始化操作完成后,通过设计适应度函数来得到同时满足距障碍物足够远与飞行路程短2个目的的轨迹;然后执行交叉与变异操作用于产生新的个体——交叉与变异发生的概率决定了种群中产生新个体的速度;通过增加删除节点的操作来避免冗余路径点的出现。实验结果表明,该方法能够得到比现有算法更安全、路程更短的平滑可行轨迹。

In order to further study on the flight safety of unmanned aerial vehicle(UAV)in autonomous flight mission,on the basis of ensuring the shortest possible flight distance,the paper proposed a path planning method based on improved genetic algorithm:the thought of keeping the UAV flight path away from obstacles was given to reduce the chance of collision caused by errors of environmental perception and flight control;the fitness function was designed to obtain a trajectory that simultaneously satisfies two goals of being far enough away from obstacles and short flight distance after the initialization of population was completed;and the crossover and mutation operations were performed to produce new individuals--the probability of their occurrence determines the rate at which new individuals are produced in the population;nodes were more deleted to avoid the occurrence of redundant path points.Experimental result showed that the proposed method could obtain a smooth and feasible trajectory that is safer and shorter in distance than existed algorithms.