一种未知环境下方向自适应的机器人避障算法

A direction adaptive obstacle avoidance algorithm for robot in unknown environment

针对机器人在未知动态环境中移动时避障难、路径规划复杂的问题,基于动态窗口法(Dynamic Window Approach,DWA)进行研究和优化,提出一种能够自适应调整方向的避障方法。首先通过障碍物信息计算可通行的间距和角度,设置临时目标点,解决了原算法会绕行密集静态障碍物区域或陷入局部最优的问题,然后在原算法的基础上新增一个用于处理动态障碍物的决策函数,将机器人与动态障碍物之间实时速度和距离的关系作为该评价函数的依据,并为该函数设置合理触发阈值,触发后通过计算选择最优的轨迹方向以完成动态避障。MATLAB实验结果证明,所提出的方法较原算法避障能力强,路径规划效率高,应对未知动态环境时有较大优势。

Aiming at the problems of difficult obstacle avoidance and complex path planning when the robot moves in an unknown dynamic environment,an obstacle avoidance method with adaptive direction adjustment is proposed based on the research and optimization of dynamic window approach(DWA).Firstly,the passable distance and angle are calculated through the obstacle information,and the temporary target point is set to solve the problem that the original algorithm will bypass the dense static obstacle area or fall into local optimization.Then,on the basis of the original algorithm,a decision function for dealing with dynamic obstacles is added.The relationship between the real-time speed and distance between the robot and dynamic obstacles is taken as the basis of the evaluation function,and a reasonable trigger threshold is set for the function.After triggering,the optimal trajectory direction is selected through calculation to complete dynamic obstacle avoidance.MATLAB experimental results show that the proposed method has stronger obstacle avoidance ability and higher path planning efficiency than the original algorithm,and has great advantages in dealing with unknown dynamic environment.