基于学习机制蚁群算法的移动机器人路径规划  

Path planning of mobile robot based on learning mechanism ant colony algorithm

作　　者：唐宏伟[1] 罗佳强 邓嘉鑫 王军权 石书琪[1] TANG Hongwei;LUO Jiaqiang;DENG Jiaxin;WANG Junquan;SHI Shuqi(Hunan Key Laboratory of Power Grid Operation and Control in Multi-source Area,Shaoyang University,Shaoyang 422000,China)

机构地区：[1]邵阳学院多电源地区电网运行与控制湖南省重点实验室,邵阳422000

出　　处：《现代制造工程》2024年第12期48-53,129,共7页Modern Manufacturing Engineering

基　　金：湖南省自然科学基金项目(2022JJ50205);湖南省教育厅科研项目(21B0682,21B0676,21C0599);湖南省科技计划项目(2016TP1023);邵阳学院研究生科研创新项目(CX2023SY083);国家级大学生创新创业训练计划项目(202210547018)。

摘　　要：针对U型障碍物环境中移动机器人路径规划问题,提出了一种学习机制蚁群算法。首先,为解决算法运行时间长的问题,引入邻域剔除,舍弃较差和对称路径;其次,为解决收敛速度慢的问题,运用禁忌策略,使蚂蚁快速逃离U型障碍物;然后,为解决路径死锁的问题,提出学习机制,不断舍弃死锁路径;最后,将该算法与其他改进算法进行仿真对比,结果表明学习机制蚁群算法相比对照组算法不仅缩短了运行时间,还提升了收敛速度,验证了该算法的优越性。A learning mechanism ant colony algorithm was proposed for the path planning problem of mobile robot in U-shaped obstacle environment.Firstly,to solve the problem of long algorithm running time,neighborhood removal was introduced to discard poor and symmetric paths.Secondly,to solve the problem of slow convergence speed,taboo strategies were applied to enable ants to quickly escape U-shaped obstacles.Then,to solve the problem of path deadlock,a learning mechanism was proposed to continuously discard deadlocked paths.Finally,a simulation comparison was conducted between the proposed algorithm and other improved algorithms.The results showed that the learning mechanism ant colony algorithm not only shortened the running time compared to the control group algorithm,but also improved the convergence speed,which verifies superiority of the algorithm.

关 键 词：路径规划 蚁群算法 邻域剔除 学习机制 移动机器人 U型障碍物 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]