改进JPS算法融合DWA的多机器人路径规划

Improved JPS algorithm fused with DWA for multi robot path planning

针对传统的跳点搜索(jump point search, JPS)算法在移动机器人路径规划时,存在路径拐点以及中间跳点过多,路径规划时间较长等问题,提出了改进的跳点搜索算法I-JPS。I-JPS算法通过改进代价函数、引入叉积公式,来剔除冗余节点、增加机器人与障碍物之间的安全距离。同时引入了动态窗口法(dynamic window approach, DWA)作局部路径规划,用于机器人临时避障和路径平滑化,并通过改进DWA提高多机器人之间的避障优先级。最后引入了多机器人协同路径规划,多机器人可以共同合作并完成复杂的任务,机器人之间还可以共享信息、协调行动,并通过分工合作来解决问题,提高任务的完成效率。最后,实验仿真结果表明改进后的算法相较于改进前的,在各方面都得到了极大的提升。

For the traditional JPS algorithm in mobile robot path planning,there are path inflection points as well as too many intermediate jump points,and the path planning time is long.This paper proposed an improved jump point search(JPS)algorithm.The I-JPS algorithm eliminated redundant nodes and increased the safe distance between the robot and obstacles by improving the cost function and introducing the fork product formula.It also introduced DWA for local path planning for temporary obstacle avoidance and path smoothing by robots,and increased the priority of obstacle avoidance between multiple robots by improving DWA.Finally,it introduced multi-robot collaborative path planning,where multiple robots could work together and complete complex tasks,and the robots could also share information,coordinate their actions,and solve problems through division of labour to improve the efficiency of task completion.Finally,the experimental simulation results show that the improved algorithm is greatly improved in all aspects compared to the pre-improved one.