摘要
针对传统拓展随机树(RRT)算法用于路径规划时存在搜索时间长、生成路径非最优、对狭窄通道环境适应性差的问题,提出了基于方向决策的快速扩展随机树路径规划(Orientation-information strategy Rapidly-Exploring Random Tree, OIS-RRT)算法。首先,通过引入基于方向变量的探索式节点,增强扩展节点导向性;其次,采用启发式采样策略,减少了冗余拓展节点;最后,利用三次B样条曲线对路径规划结果进行平滑处理。由对比实验可知,改进算法有利于提升收敛速度并增强算法对于复杂环境的适应能力,避免陷入局部最优。仿真结果表明,改进算法在提高算法效率的同时能够实现更高质量的路径规划。
The traditional rapid expanding random tree(RRT) algorithm exists many disadvantages in path planning, such as long search time, non-optimal path generation and low sensitivity to narrow channels. Therefore, a mobile robot path planning based on the orientation-information strategy RRT(OIS-RRT) algorithm is proposed. Firstly, the orientations of extended nodes were enhanced by introducing exploratory nodes based on directional variables;secondly, a heuristic sampling strategy was adopted to reduce redundant extended nodes;finally, the results of path planning were smoothed by using a cubic B-spline curve. It can be seen from the comparative experiments that the improved algorithm is beneficial to increase the convergence speed and enhance the algorithm’s adaptability to complex environments, and avoid falling into local optimum. The simulation results show that the algorithm in this paper can achieve higher quality path planning while improving the efficiency of the algorithm.
作者
刘想德
何翔鹏
胡勇
胡小林
LIU Xiang-de;He Xiang-peng;HU Yong;HU Xiao-lin(Research and Development Center of Information Accessibility Engineering and Robotics,Chongqing University of Posts and Teleco mmunications,Chongqing 400065,China;Chongqing Innovation Center of Industrial Big-Data Co.Ltd.,Chongqing 400000,China)
出处
《计算机仿真》
北大核心
2022年第6期444-448,495,共6页
Computer Simulation
基金
国家自然科学基金(51905065)
重庆市北碚区科学技术局技术创新与应用示范项目(2020-5)。
关键词
路径规划
方向决策
启发式采样
Path planning
Orientation strategy
Heuristic sampling
