基于改进RRT算法的工业机器人路径规划研究  

Industrial Robot Path Planning Based on Enhanced RRT Algorithm

作　　者：仪凌霄 黄永裕 潘斯宁 张俊强 黄广树 YI Lingxiao;HUANG Yongyu;PAN Sining;ZHANG Junqiang;HUANG Guangshu(College of Artificial Intelligence,Hezhou University,Hezhou 542899,China;College of Mechanical and Electrical Engineering,Guilin University of Electronic Technology,Guilin 541004,China)

机构地区：[1]贺州学院人工智能学院,广西贺州542899 [2]桂林电子科技大学机电工程学院,广西桂林541004

出　　处：《机械工程师》2025年第2期14-20,25,共8页Mechanical Engineer

基　　金：国家自然科学基金(52065018);广西科技基地与人才专项(AD20297043);广西自然科学基金(2022GXNSFB A035597);广西高校中青年教师基础科研能力提升项目(2022KY0706);贺州市科学研究与技术开发计划项目(贺科技202228)。

摘　　要：针对快速扩展随机树(Rapidly-exploring Random Trees,RRT)算法在工业机器人复杂工况中进行避障等路径规划中存在的运算效率低、收敛速度慢、搜索空间盲目性等问题,提出一种基于RRT算法的工业机器人路径规划改进算法。应用标准D-H(Denavit-Hartenberg)参数法建立六自由度工业机器人的运动学模型,并通过数值分析研究了机器人的正运动学和逆运动学模型,使用机器人工具箱验证了运动学模型的可行性。以机器人运动路径成本和运算时间作为优化目标,依次对改进RRT算法和RRT、Bi-RRT(Bidirectional-RRT)算法进行不同维度的仿真分析和比较。研究结果表明,改进RRT算法优于其他两种算法,将改进RRT算法应用于机器人避障和路径规划中具有良好的效果。Aiming at the problems of low computational efficiency,slow convergence speed,and blindness of search space that exist in the Rapidly-exploring Random Trees(RRT)algorithm for obstacle avoidance and other path planning of industrial robots in complex working conditions,this paper puts forward a kind of improved algorithm for path planning of industrial robots based on the RRT algorithm.The standard D-H(Denavit-Hartenberg)parameter method is applied to establish the kinematic model of a six-degree-of-freedom industrial robot,and the forward kinematics and inverse kinematics of the robot are investigated by numerical analysis,and the feasibility of the kinematic model is verified by using the robot toolbox.Taking the robot motion path cost and computing time as the optimization objectives,the improved RRT algorithm and the RRT and Bi-RRT(Bidirectional-RRT)algorithms are sequentially simulated and compared in different dimensions.The results show that the modified RRT algorithm outperforms the previous two methods,and it can be successfully used for robot path planning and obstacle avoidance.

关 键 词：工业机器人 运动学分析 路径规划 改进RRT算法 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]