融合人工势场和Informed-RRT^(*)算法的机械臂自适应路径规划  

Adaptive path planning of manipulators combining Informed-RRT^(*)with artificial potential field

作　　者：贾浩铎 房立金[2] 王怀震 JIA Haoduo;FANG Lijin;WANG Huaizhen(School of Information Science and Engineering,Northeast University,Shenyang 110819,China;Faculty of Robot Science and Engineering,Northeast University,Shenyang 110819,China;Shandong New Generation Information Industry Technology Research Institute,Inspur,Jinan 250101,China)

机构地区：[1]东北大学信息学院,辽宁沈阳110819 [2]东北大学机器人学院,辽宁沈阳110819 [3]浪潮集团山东新一代信息产业技术研究院有限公司,山东济南250101

出　　处：《计算机集成制造系统》2025年第4期1179-1189,共11页Computer Integrated Manufacturing Systems

基　　金：辽宁省应用基础研究计划资助项目(2022JH2/101300202)。

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。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.

关 键 词：Informed-RRT^(*)算法 人工势场法 路径规划 动态场景 

分 类 号：TP241[自动化与计算机技术—检测技术与自动化装置]