基于ACO-RRT算法的移动机械臂避障路径规划  

Obstacle avoidance planning of mobile manipulator based on ACO-RRT

作　　者：陈春朝[1] 刘义 王耿[1] CHEN Chunchao;LIU Yi;WANG Geng(School of Mechanical and Dynamic Engineering,Henan Polytechnic University,Jiaozuo 454000,Henan,China)

机构地区：[1]河南理工大学机械与动力工程学院,河南焦作454000

出　　处：《河南理工大学学报（自然科学版）》2023年第3期95-102,共8页Journal of Henan Polytechnic University（Natural Science）

基　　金：国家自然科学基金资助项目(51505133);河南省高等学校重点科研项目(19A460004);河南省科技攻关项目(202102-210082);河南理工大学博士基金项目(B2019-48)。

摘　　要：在充满障碍物的环境里,基于随机采样的RRT算法和RRT^(*)算法进行移动机械臂避障路径规划时常存在路径代价高和求解速度慢等问题。为解决以上问题,首先将ACO蚁群算法中的启发式思想融入RRT算法形成ACO-RRT算法,在不同使用场景引入不同启发式修改采样策略解决最短路径问题;然后通过定义效用函数权衡探索函数与开发函数分配率,使得在效用函数增加的地方采样更加频繁;最后在相同的障碍环境中评估ACO-RRT、RRT和RRT^(*)3种算法的迭代次数、路径代价和求解速度等性能指标。实验结果表明,ACO-RRT算法解决了RRT算法和RRT^(*)算法存在的问题,并且性能优于目前较为成熟的RRT算法和RRT^(*)算法。In an environment full of obstacles,the path planning of mobile manipulators based on random sampling RRT algorithm and RRT^(*)algorithm usually have problems such as the high cost of path and the slow speed of solution.In order to solve the above problems of RRT and RRT^(*),ACO heuristics were inte-grated into RRT algorithm to form ACO-RRT algorithm,and were set with different values in different sce-narios to solve the shortest path problem by modifying the sampling strategy in this paper.Then,a new util-ity function was defined to balance the distribution rate of exploration function and development function,the utility function made the sampling more frequent where increased.Finally,by comparing the number of iterations,path cost and solution speed,ACO-RRT,RRT and RRT^(*)algorithms were evaluated in the same obstacle environment.It was shown that the ACO-RRT algorithm solved the problems of the RRT algorithm and the RRT^(*)algorithm,and its performance was better than the current mature RRT and RRT^(*).

关 键 词：移动机械臂 避障路径规划 RRT算法 RRT*算法 ACO算法 效用函数 

分 类 号：TP241[自动化与计算机技术—检测技术与自动化装置]