基于动态窗口法与人工势场法融合的差速机器人路径规划算法

A differential robot path planning algorithm combining the dynamic window algorithm and the artificial potential field

目的:原始动态窗口法在路径规划时容易出现路径震荡现象,为此本文设计了一种与人工势场法融合的改进动态窗口法,以减少路径震荡。方法:将原始动态窗口法中的线性障碍物距离评价函数,改进为人工势场法中非线性障碍物势场函数。通过MATLAB仿真和硬件实验与原始算法进行对比,验证算法的可行性。结果:本文改进的动态窗口法与原始算法相比,减少了机器人运行时的震荡,路径更加平滑。改进算法在仿真中场景中路径长度减少2.5%,运行时间减少40.9%;硬件实验中,路径长度减少8.8%,运行时间减少13.9%。结论:改进的动态窗口法生成的路径具有更好的平滑性,同时路径长度和运行时间较短。

Aims:An improved dynamic window method combined with the artificial potential field method was designed to reduce the path oscillations of the traditional algorithm.Methods:The linear obstacle distance evaluation function in the traditional DWA was improved to be a nonlinear obstacle potential field function in the APF.The feasibility of the algorithm was verified as compared with the traditional algorithm through the MATLAB simulation and experiment.Results:Compared with the original algorithm,the improved dynamic window method reduced the oscillation;and the path was smoother.In the simulation,the improved algorithm reduced the path by 2.5%and the running time by 40.9%.In the hardware experiment,the path was reduced by 8.8%;and the running time was reduced by 13.9%.Conclusions:The path generated by the improved dynamic window method has better smoothness and shorter path length and running time.