基于改进HybridA^(*)的非完整约束机器人路径规划算法研究

Path Planning of Robot with Nonholonomic Constraint based on Improved Hybrid A^(*)algorithm

为了实现单舵轮AGV在物流场景下的精准自主导航,针对基于HybridA*算法搜索的路径容易贴近障碍物的缺陷和算法在路径平滑后可能与障碍物冲突的问题,本文提出一种基于改进HybridA*的非完整约束轮式移动机器人路径规划方法。对于路径搜索部分,将距离场地图引入启发式函数,以提高搜索效率,并保持搜索路径远离障碍物;对于路径平滑部分,将优化问题转化为二次规划问题,通过边界约束保证避障效果。在ROS平台上仿真实验的结果表明,该路径规划方法可有效用于非完整约束系统,为单舵轮AGV自主导航的准确运动奠定了路径智能规划的基础。

In order to achieve precise autonomous driving of the single steering wheel automated guided vehicle(AGV)in logistics scene,this paper proposes a path planning method for wheeled mobile robots(WMRs)with nonholonomic constraint based on improved Hybrid A*algorithm,so that the problems of original Hybrid A*algorithm on obstacles avoidance could be solved.For the path-search part,a distance field is introduced into the heuristic function,which could improve the searching efficiency and keep the search path away from obstacles.For the path-smoothing part,a quadratic programming problem is described to guarantee the obstacle avoidance through boundary constraints.The results of simulated experiments on ROS show that the method proposed could be effectively applied to nonholonomic constrained robots,and could lay the foundation for the accurate movement of single steering wheel AGV on autonomous navigation.