基于BD-RRT算法的动态路径规划研究

Research on Dynamic Path Programming Based on BD-RRT Algorithm

针对Bi-RRT算法规划路径不平滑、路径生成冗余、规划时间长和动态环境不适用等问题,展开了基于BD-RRT(bidirectional-dynamic-RRT)算法的动态路径规划研究,以解决实际场景中障碍物变化时的避障问题。在路径生长方面,提出动态步长增长策略提高对障碍物附近点的采样精度,减少采样点以增强算法的探索能力;在环境发生改变时,引入动态RRT算法,可以修剪首次规划路径中有碰撞的节点,在无碰撞的节点的基础上继续重规划;采取B样条曲线平滑生成的路径以达到减少能耗的目的。仿真结果表明,在静态环境中,BD-RRT算法较其它传统算法能生成更简短有效的路径,路径规划的时间更短,生成的路径更加平滑,且平均路径规划长度与时间分别为Bi-RRT算法的21.1%和95.3%,证明了改进的有效性;在动态环境中,BD-RRT算法的平均重规划时间和路径长度都短于Extended-RRT,仅为该算法的5.2%和95.8%,有着更好的动态环境实用性。

Aiming at the problems of unsmooth path planning,redundant path generation,long planning time and inapplicable dynamic environment of Bi-RRT algorithm,the research of dynamic path planning based on BD-RRT(bidirectional-dynamic-RRT)algorithm is carried out to solve the obstacle avoidance problem when obstacles change in actual scenes.In terms of path growth,a dynamic step growth strategy is proposed to improve the sampling accuracy of points near obstacles and reduce sampling points to enhance the exploration ability of the algorithm.When the environment changes,the dynamic RRT algorithm is introduced,which can prune the nodes with collisions in the first planning path,and continue to replan on the basis of collision-free nodes.B-spline curves are used to smooth the generated path to reduce energy consumption.The simulation results show that in the static environment,the BD-RRT algorithm can generate shorter and more effective paths than other traditional algorithms,the path planning time is shorter,the generated paths are smoother,and the average path planning length and time are 21.1%and 95.3%of the Bi-RRT algorithm,respectively,which proves the effectiveness of the improvement.In the dynamic environment,the average replanning time and path length of the BD-RRT algorithm are shorter than that of Extended-RRT,which is only 5.2%and 95.8%of the algorithm,which has better practicability in the dynamic environment.