机器人在已知三维自然环境中的路径规划算法

Path planning algorithm for robot in known three-dimensional natural environment

研究了移动机器人在三维环境的路径规划问题,针对该问题中存在环境适应性和全局性差的不足,对人工势场法进行改进,提出了一种新的路径规划的算法。该算法首先对已知的三维自然环境进行栅格化,建立栅格运行费用的评估模型,计算每个栅格的运行费用;然后依据栅格的运行费用建立斥力场,以目标点为中心的建立引力场,同时提出解决局部最小值的问题的方法;最后将两者合力的方向作为移动机器人在该点的路径走向,规划出一条从起始点到目标点的运行费用较低的路径。仿真实验结果表明,该算法能有效降低运行费用,适应性和全局性好,适合应用于移动机器人在三维自然环境中的路径规划。

A path planning algorithm for robot is studied to improve the shortcomings of mobile robot path planning in three-dimensional natural environment.The artificial potential field（APF） is improved and a path planning algorithm is proposed.Firstly,the known three-dimensional environment is rasterized and running costs evaluation model for each grid is established.Secondly,repulsion field is established based on the grid running costs and gravitational field with target point at its centre,also a solution to the local minimum problem is proposed.Finally,we use the direction of the resultant force to plan a path with lower costs from starting point to the target.Simulation results show that the algorithm has lower operating costs,good adaptability and global property.