摘要
移动机器人在行驶过程中需要躲避静态和动态双重障碍物,针对传统DWA(dynamic window approach,DWA)算法在选择路径时,受静态和动态双重障碍物的影响,容易造成算法运算缓慢、易陷入局部最优解、运动状态不稳定的问题,提出了一种融合多物理因素的DWA局部路径规划算法。在传统DWA算法基础上,设置障碍物阈值对距离评价函数进行改进,减少机器人受较远障碍物的影响,加快判定速度;采用自适应机制对算法的安全系数进行动态调节,解决局部最优解及路径不平滑的问题;通过改进归一化函数,减少各个物理量之间的差异,使机器人能够快速选择出最优速度,保证机器人运动状态的稳定;最后,通过两组不同的仿真及模拟场地实验进行对比验证,实验结果证明了改进后的算法相较于原始算法在判定速度、路径平滑度、路径安全性、运动稳定性方面均有明显提升。
Mobile robots need to avoid both static and dynamic obstacles while in motion.Traditional DWA(dynamic window approach)algorithm is affected by both static and dynamic obstacles when selecting a path,easily causing slow operation,local optimal solution,and unstable motion state.This paper proposes a DWA local path planning algorithm based on multiple physical factors.According to the traditional DWA algorithm,an obstacle threshold is set to improve the distance evaluation function,reducing the impact of distant obstacles and improving the judgment of speed.The adaptive mechanism is employed to dynamically adjust the safety factor of the algorithm to address local optimal solution and unsmooth path.By improving the normalization function,the difference between each physical quantity is reduced,so that the robot quickly decides the optimal speed and ensures the stability of its motion state.Finally,two sets of different simulation and site experiments are compared and verified.Results show the improved algorithm significantly improves the judgment of speed,path smoothness,path safety and motion stability.
作者
王丰
陈洋
王志军
于浩
WANG Feng;CHEN Yang;WANG Zhijun;YU HAO(College of Mechanical Engineering,North China University of Technology,Tangshan 063210,China)
出处
《重庆理工大学学报(自然科学)》
北大核心
2025年第7期142-148,共7页
Journal of Chongqing University of Technology:Natural Science
基金
唐山市科技创新团队培养计划项目(21130208D)。
关键词
移动机器人
改进DWA算法
局部路径规划
静动态避障
mobile robot
enhanced DWA algorithm
local path planning
static and dynamic obstacle avoidance
