不确定环境下基于改进萤火虫算法的地面自主车辆全局路径规划方法

Global Path Planning for ALV Based on Improved Glowworm Swarm Optimization Under Uncertain Environment

针对地面自主车辆的特点,提出了一种基于改进萤火虫算法(Glowworm Swarm Optimization,GSO)的路径规划方法.首先利用GSO覆盖多个局部最优解的能力,一次生成多条规划路径;然后提出两种路径切换算法,分别用于调优和脱困.在通过路径交叉点时,调优切换算法对交叉路径进行重新评估并切换到较优路径,最终达到实际行驶路径的最优化.在遇到环境发生改变时,脱困切换算法通过启发式搜索快速切换到适当路径,重用了原搜索结果,避免了二次规划.通过大量的仿真实验及实际试用,证明了所提方法的可行性和有效性.

According to the characteristics of autonomous land vehicle, a global path planning method based on improved glowworm swarm optimization （GSO） is proposed. Firstly, more than one path is generated with GSO which covers multiple local optima. Then two path switching algorithms are proposed, of which one alms at optimization and the other aims at rescue. When the cross point is passed through, the optimization switching algorithm revaluates the paths, switches to the optimum path, and ultimately attains optimal actual travel route. When the environment changes, the rescue switching algorithm switches to the appropriate path by heuristic search, which reuses the original search results, avoiding the secondary planning. Many simulation experiments and actual trial show that the proposed method is feasible and effective.