摘要： A^*算法通过启发信息指引搜索方向，被广泛应用于移动机器人的路径规划，但其规划出的搜索路径存在冗余节点且与障碍物相近，无法满足动态避障需求。对标准A^*算法进行改进，设计安全A^*算法并融合动态窗口法进行路径规划。定义安全距离因子引入A^*算法的启发函数中，提高算法规划路径的安全性，同时采用平面结构法对算法规划得到的路径进行优化，根据相邻节点与障碍物之间的位置关系判断该相邻节点间是否存在障碍物，由此减少路径拐点数，提高路径平滑度。由于当移动机器人处于未知环境时，仅靠A^*算法不能避开障碍物到达目标点，因此借助动态窗口法的局部避障功能。通过安全A^*算法规划全局最优路径节点坐标，设计融合子函数改进动态窗口法的评价函数，解决动态窗口法易陷入局部最优的问题。实验结果表明，在复杂环境中，该方法通过融合安全A^*算法和动态窗口法，能够确保在安全路径基础上实时随机避障，使机器人安全到达终点。

关键词: A^*算法, 安全距离因子, 平面结构法, 动态窗口法, 路径规划

Abstract: The A^* algorithm relies on heuristic information to guide the search direction and is widely used for path planning of mobile robots.However, the search path of this algorithm has redundant nodes that are similar to obstacles, which prevent the algorithm from meeting the requirements of dynamic obstacle avoidance.This study aims to improve the standard A^* algorithm by designing a safe A^* algorithm and integrating the Dynamic Window Approach(DWA) for path planning.Particularly, the safety distance factor is defined and introduced into the heuristic function of the A^* algorithm to improve the safety of the path planned by the algorithm.Simultaneously, the plane structure method is used to optimize the path planned by the algorithm.The relationship between the positions of adjacent nodes and obstacles is used to determine whether obstacles exist between adjacent nodes to reduce the number of inflection points of the path and improve the smoothness of the path.When the mobile robot is in an unknown environment, the A^* algorithm alone cannot prevent the robot from avoiding obstacles on its way to reaching the target point;therefore, the local obstacle avoidance function of the DWA is used.The global optimal path node coordinates are planned using the safe A^* algorithm.In addition, the fusion sub-function is designed to improve the evaluation function of the DWA to solve the tendency of the DWA to easily converge to local optima.The experimental results show that in a complex environment, integration of the safe A^* algorithm and the DWA ensures real-time random obstacle avoidance based on the safe path, thereby enabling the robot to reach the end point safely.

Key words: A^* algorithm, safety distance factor, plane structure method, Dynamic Window Approach(DWA), path planning

中图分类号: 

• TP391