基于蚁群算法的移动机器人路径规划

摘要

研究移动机器人在已知静态环境下路径规划问题,在避障环境下寻求最优路径.针对蚁群算法搜索时间长、易陷入局部最优等缺陷,导致实时处理困难,且路径准确度低、可跟踪性差不能直接用于机器人.为解决上述问题,首先提取环境的平面几何信息,建立了简单有效地搜索模型.可通过引入终点距离与方向的启发函数、阶梯式伪随机的结点转移规则,引导蚁群有目的的进行搜索;改进信息素更新策略,利用一种奖惩机制以增强蚁群对尽可能好的解的识别能力.并考虑障碍物对路径的影响,运用人工势场法对全局最优路径的结点进行平滑.进行仿真的结果表明,提高了路径的安全性和可跟踪性.证明了改进方法可以有效地找出最优可行路径.%Ant Colony Algorithm needs long searching time, and easily falls into local optimal solution , resulting in the defects of poor real-time processing, and low. E accuracy and traceability. To solve the problem, a simple efficiently search model was established based on the plane geometry information extracted. The heuristic function of the end's distance and direction, and the pseudo-random ladder node transition rule were introduced to guide ant colony to search with the purpose. Pheromone update strategy was improved with an award and penalty mechanism to enhance the ability of ant colony to identify the best possible solution. Finally, the optimal path' s node was smoothed by artificial potential field. The experimental results show that the method suggested can effectively search out the optimal feasible path.