移动机器人平滑JPS路径规划与轨迹优化方法

针对目前路径规划方法存在的平滑性和效率问题,在JPS算法基础上提出了兼顾平滑性与搜索效率的路径规划方法,并利用多项式进行了轨迹优化。首先,提出2个优化目标对路径序列进行优化处理;然后,对JPS搜索规则进行改进,得到更多有价值的路径,并对每条路径进行平滑处理,再以一定规则进行选择;最后,使用多段高阶多项式对所得路径进行轨迹优化,研究时间分配问题,从而加快迭代效率。通过仿真实验和与其他算法的对比证明了本文方法的可行性和有效性。结果表明,在不同障碍物密度环境下,本文路径规划方法得到了平滑性良好的路径,相对平滑后处理JPS,长度减少了0.48%~1.80%,总转折角减少了16.93%~52.75%,利用余弦函数进行时间分配加快了轨迹优化的迭代效率,通过实验验证得到了良好的效果。

Smooth JPS Path Planning and Trajectory Optimization Method of Mobile Robot

Aiming at the problems of smoothness and efficiency in current path planning methods, based on jump point search (JPS), a path planning method considering both smoothness and search efficiency was proposed, and the trajectory was optimized by polynomial. Firstly, in order to improve the smoothness of the path, two optimization objectives were proposed to optimize the path sequence, that was, each corner of the path was on the vertex of the obstacle grid, and the obstacles that contacted with the path corner were located on the side with the angle less than 180 degrees. Then, the search rules of JPS were improved to get more valuable paths, and each path was smoothed, and then the path selection was made by weighing the length and angle with certain rules. Finally, polynomial was used to optimize the path, and the time allocation of the trajectory optimization method was studied to speed up the iterative efficiency. The feasibility and effectiveness of this method were proved by simulation and comparison with other algorithms. The results showed that the method had the advantages of high efficiency of JPS algorithm, and effectively solved the problems of redundant points and frequent turning points in path planning of JPS algorithm. In different density environments, compared with the smoothed JPS algorithm, the path planning method reduced the path length by 0.48%~1.80%, and the cumulative turning angle was decreased by 16.93%~52.75%. In the large environment, the proposed method had similar smoothness to Lazy-Theta*, but it had higher search efficiency. The cosine function was used to allocate time to accelerate the iterative efficiency of trajectory optimization, and good results were obtained through experiments.