基于改进蝙蝠算法的无人机路径规划研究

Research on UAV Path Planning Based on Improved Bat Algorithm

针对蝙蝠算法在处理无人机路线规划问题时出现的前期迭代缓慢、容易进入局部最优等情况。论文将遗传因子算法与蝙蝠算法相结合,提出一种融合遗传因子的蝙蝠算法(Genetic-Bat Algorithm,GBA)。算法首先引入遗传算法的交叉和变异操作进行选择融合,提高前期迭代收敛速度;然后在蝙蝠算法中使用分阶段局部搜索,增加对最优解局部域的检索;最后加入删除操作,以减少路径冗余度。通过建立地图模型的仿真,结果表明该算法在路径规划和迭代速度上和其他算法相比能够实现快速收敛和快速迭代,不易陷入局部最优解,选择的路径更优。

The bat algorithm is slow in the early iteration and easy to enter the local optimum when dealing with UAV route planning.In this paper,a Genetic-Bat Algorithm(GBA)is proposed by combining genetic factor algorithm with bat algorithm.First⁃ly,crossover and mutation operations of genetic algorithm are introduced for selective fusion to improve the convergence speed of earlier iteration.Then in the bat algorithm,the local search by stages is used to increase the search of the optimal solution.Finally,delete operations are added to reduce path redundancy.Through the simulation of map model,the results show that compared with other algorithms in path planning and iteration speed,the proposed algorithm can achieve fast convergence and fast iteration,and it is not easy to fall into the local optimal solution,so the selected path is better.