改进的RRT路径规划算法  

Improved RRT path planning algorithm

作　　者：刘恩海 高文斌[1,2] 孔瑞平 刘贝野 董瑶 陈媛媛 LIU En-hai;GAO Wen-bin;KONG Rui-ping;LIU Bei-ye;DONG Yao;CHEN Yuan-yuan(School of Artificial Intelligence,Hebei University of Technology,Tianjin 300401,China;Hebei Provincial Key Laboratory of Big Data Computing,Hebei University of Technology,Tianjin 300401,China;Department of Intelligent Technology,Tianjin Polytechnic College,Tianjin 300400,China)

机构地区：[1]河北工业大学人工智能与数据科学学院,天津300401 [2]河北工业大学河北省大数据计算重点实验室,天津300401 [3]天津工业职业学院智能技术系,天津300400

出　　处：《计算机工程与设计》2019年第8期2253-2258,共6页Computer Engineering and Design

基　　金：天津市科技计划基金项目(14ZCDGSF00124);天津市基础研究计划基金项目(17JCTPJC55400、17JCTPJC55600)

摘　　要：对全局路径规划算法中的快速扩展随机树(RRT)算法进行深入的研究,针对基本RRT算法随机性强、搜索没有偏向性、得到的路径不一定为最优路径等缺点,提出一种改进的RRT优化算法,通过改进随机数生长方式的角度对原有算法思路进行改进,引导随机数向着目标点方向生长,与此同时移动机器人可以根据周围环境信息及时做出调整,使随机树向更高质量生长。大量仿真结果表明,改进的RRT算法具有一定的可行性与有效性,能高效引导随机树朝目标点方向高质量地生长,规划的路径尽可能接近最优路径,有效缩短了路径规划时间。This research based on rapidly-exploring random trees (RRT) algorithm in global path planning to improve shortcomings of traditional RRT,such as random-inclinding,lacks of preference in searching and local-or-global optimum problem.Owing to these shortcomings,an improved RRT algorithm was proposed,which leaded random number to grow and get closer to the target by improving the angle of the growth mode of random number,and meanwhile the robot made adjustments according to environment,which enabled the random tree to have better qualities.Numerous simulation experimental results show that this improved RRT algorithm is feasible and effective and it can lead the random tree to grow and get closer to the target effectively,its result is closed to global optimum with less time consumed.

关 键 词：路径规划 RRT算法 目标偏向 路径优化 人工势场算法 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]