Abstract: For the problems of long planning time,low iteration efficiency and poor applicability of change environment in the path planning of Informed-RRT* algorithm,an adaptive path planning of manipulators combining Informed-RRT* with artificial potential field was proposed.In the direction of path growth,a probability adaptive target bias strategy was presented,which constructed the judgment region to generate the bias probability,and combined with artificial potential field constraints to limit the randomness of path direction selection.In the path expansion,a global adaptive step size method was proposed,which adjusted the step size according to the spatial position of the sampling points in the artificial potential field to improve the path exploration ability and shorten the planning time.In the path iteration,the Position Guided Function was used to guide the generation of iteration points,and the path optimization iteration was carried out efficiently.After changing the scene,the old tree information was retained,and the artificial potential field method was used to replan the path.By reselecting the target point,the local optimal trap was jumped out,and the applicability of the algorithm in the dynamic scene was enhanced.The simulation results showed that the proposed algorithm could improve the speed of path planning by 51.59% and reduce the optimal path length by 8.03% compared with Informed RRT* algorithm,so it had stronger adaptability when the environment changed.

Key words: Informed-RRT* algorithm, artificial potential field, path planning, dynamic scene

摘要： 针对Informed-RRT*算法存在规划用时长、迭代效率低、动态场景不适用的问题,提出一种融合人工势场和Informed-RRT*算法的机械臂自适应路径规划算法。在路径生长方向上,提出一种概率自适应的目标偏置策略,构造判定区域生成偏置概率,结合人工势场约束,限制路径方向选择的随机性;在路径扩展中,提出一种全局自适应步长方法,根据采样点在人工势场中的空间位置调整步长,提高路径探索能力,缩短规划用时;在路径迭代中,采用位置函数引导迭代点生成,高效地进行路径优化迭代;在场景变动后,保留旧树信息,利用人工势场方法进行路径重规划,通过重选目标点跳出局部最优陷阱,增强算法在动态场景的适用性。仿真结果表明,与Informed-RRT*算法相比,所提算法在路径规划速度方面提高51.59%,最优路径长度减少8.03%,在环境变化时具有更强的适应性。

关键词: Informed-RRT*算法, 人工势场法, 路径规划, 动态场景