基于动态离散势场的迷宫机器人路径规划

Path Planning for Maze Robot Based on Dynamic Discrete Potential Field

针对未知环境下的机器人迷宫求解问题,提出一种动态离散势场路径规划算法。为提高路径优化性能,采用引入边界节点的栅格法建立模型,在各栅格的边界节点处定义障碍物状态和势场的数值大小,通过计算可连通相邻节点的累计代价值完成势场的构造。为提高寻优速度,随着环境信息的更新动态改变势场分布,沿势场下降最快方向获得实时重规划路径,引导机器人向目标运动,通过预规划路径的访问状态判断路径是否收敛,避免无用栅格的扩展。仿真实验结果表明,应用该算法可使机器人在复杂未知的迷宫环境中快速、高效地规划出一条折线少、转折角度小的优化路径。

Aiming at robot solving maze problem in the unknown environment, a path planning algorithm based on dynamic discretely potential is proposed. In order to improve the performance of path optimization, the algorithm uses grids method which including boundary node to model, and by accumulating the values of connected nodes which mean the grid＇s status of obstacles to build potential field. Furthermore, to increase search rate, the algorithm dynamically changes potential field as the environment information updated, and obtains the direction which potential values fall to guide the robot moving toward target. The convergence of path is judged by its visit status to avoid extension of useless grids. Simulation experimtentaI results show that the robot can find out smoothing optimal path quickly and efficiently by using the algorithm in the complex and unknown maze environment.