融合RRT-Connect和DWA算法的室内移动机器人单目标点导航任务研究  

Research on single-target point navigation task of indoor mobilerobot integrating RRT-Connect and DWA algorithms

作　　者：王冠强 张驰洲 陈明松 蔺永诚[1,2] 邹奋扬 王秋 吴敏杰 曾维栋 WANG Guanqiang;ZHANG Chizhou;CHEN Mingsong;LIN Yongcheng;ZOU Fenyang;WANG Qiu;WU Minjie;ZENG Weidong(School of Mechanical and Electrical Engineering,Central South University,Changsha 410083,China;State Key Laboratory of Precision Manufacturing for Extreme Service Performance,Changsha 410083,China;Light Alloy Research Institute,Central South University,Changsha 410083,China)

机构地区：[1]中南大学机电工程学院,湖南长沙410083 [2]极端服役性能精准制造全国重点实验室,湖南长沙410083 [3]中南大学轻合金研究院,湖南长沙410083

出　　处：《中南大学学报（自然科学版）》2023年第11期4326-4337,共12页Journal of Central South University:Science and Technology

基　　金：国家重点研发计划项目(2022YFB3706902);湖南省研究生科研创新项目(CX20220264)

摘　　要：针对室内复杂环境下移动机器人单目标点导航问题,提出一种基于改进双向快速扩展随机树(RRTConnect)算法与改进动态窗口法(dynamic window approach,DWA)的单目标点导航算法。首先,结合人工势场法与RRT-Connect算法设计全局规划器,并引入范围限定函数及两棵双向RRT随机树进行扩展,优化采样和路径扩展过程,同时对全局规划器生成的路径进行平滑处理;然后,改进DWA中的评价函数,以目标点距离函数代替航向角函数以加快算法收敛并避免角度反复调整,同时实现机器人的高效动态避障;最后,以全局规划器的路径点为引导,对局部规划器进行融合以完成单目标点导航任务。研究结果表明:本文所提出的规划器能够实现机器人在复杂环境下的自主路径规划和避障;相比于其他方法,本文所提出的方法效率高,规划路径短。Aiming at the single-target point navigation problem of mobile robots in complex indoor environments,a single-target point navigation algorithm was proposed based on the improved bidirectional rapidly expanding random tree(RRT-Connect)and the improved dynamic window approach(DWA)methods.Firstly,a global planner was designed by combining the artificial potential field method and the RRT-Connect algorithm,and a range limiting function and two bidirectional RRT random trees were introduced for expansion,optimizing the sampling and path expansion process.Furthermore,the paths generated by the global planner were also smoothed during the sampling and path expansion.Secondly,the heading angle function in DWA was replaced by the distance function of the target point to speed up the algorithm convergence and avoid repeated angle adjustment,and at the same time to realize the efficient dynamic obstacle avoidance of the robot.Finally,guided by the waypoints of the global planner,the local planners were fused to complete the single-target point navigation task.The results show that the planner proposed in this paper can realize autonomous path planning and obstacle avoidance of robots in complex environments.Compared with other methods,the proposed method has high efficiency and short planning paths.

关 键 词：路径规划 动态避障 RRT-Connect DWA 人工势场法 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]