基于改进融合蚁群算法的移动机器人路径规划

Path planning of mobile robot based on improved fusion ant-colony algorithm

针对蚁群算法搜索盲目、全局性差及收敛慢等缺点,文中提出了一种改进融合蚁群算法。采用扩展后的Dijkstra算法对转移概率进行改进,扩展了搜索范围,并引入终点信息对初始信息素改进,增强了初期路径搜索的导向性;在启发函数中加入自适应因子与终点信息,提高算法的自适应性与收敛速度。同时,引入自适应挥发因子对信息素更新优化,增强全局搜索能力。为避免与障碍物边缘触碰,将改进的蚁群算法与斥力场融合,并采用B样条曲线对路径进行优化,得到连续平滑的最佳移动路径。通过试验将改进的融合蚁群算法与其他蚁群算法在不同环境下进行比较,结果表明:该算法在路径长度、收敛速度、触碰障碍物边缘次数和转弯次数方面都优于其他蚁群算法。

Since the traditional ant-colony algorithms have such shortcomings as blind search,poor global exploration,and slow convergence,in this article an improved fusion ant-colony algorithm is proposed.The expanded Dijkstra algorithm is employed to enhance the transition probability,and thus the search scope is expanded.The endpoint information is introduced to modify the initial pheromone levels,in order to enhance the guidance of early path exploration.Both the adaptive factor and the endpoint information are incorporated into the heuristic function,so as to improve the algorithm's adaptability and convergence speed.Simultaneously,the adaptive evaporation factor is introduced to optimize the pheromone levels,for better capability of global exploration.In order to avoid contact with obstacle edges,the improved ant-colony algorithm is combined with a repulsive field,and the B-spline curve is used for path optimization;as a result,an optimal path is obtained,which is continuous and smooth.Through a series of experiment,the improved fusion ant-colony algorithm is compared with other ant-colony algorithms in various environments.The results indicate that this algorithm outperforms its counterparts in terms of path length,convergence speed,frequency of contact with obstacle edges,and number of turns.