第17卷第4期 智能系统学报 Vol.17 No.4 2022年7月 CAAI Transactions on Intelligent Systems Jul.2022 D0:10.11992/tis.202108020 网络出版地址：https:/kns.cnki.net/kcms/detail/23.1538.TP.20220518.0930.002.html 分组教学蚁群算法改进及其在机器人路径规划中应用 蒲兴成2，宋欣琳 (1.重庆邮电大学计算机科学与技术学院，重庆400065,2.重庆邮电大学理学院，重庆400065) 摘要：针对蚁群算法收敛速度慢、易陷入局部最优问题，提出一种基于分组教学优化改进蚁群算法。该算法 从3个角度对蚁群算法进行改进。首先，利用分组教学优化算法改进蚁群算法适应度函数，提高算法全局求解 能力。同时，引进一种新的回退策略，通过该策略处理U型障碍死锁问题，确保算法求解可行性。其次，采用 一种新的动态信息素更新策略，滚动更新每轮迭代后路径信息素值，避免算法陷入局部最优。最后，引入路径 简化算子，将冗余角简化为直线路径，缩短路径长度。仿真实验证明改进算法能有效提高移动机器人路径规划 收敛速度和精度。 关键词：改进蚁群算法：分组教学优化：路径规划：移动机器人：信息素更新：启发式函数：路径简化：回退策略 中图分类号：TP273文献标志码：A文章编号：1673-4785(2022)04-0764-08 中文引用格式：蒲兴成，宋欣琳.分组教学蚊群算法改进及其在机器人路径规中应用八.智能系统学报，2022,17(4)： 764-771. 英文引用格式：PU Xingcheng,.SONG Xinlin.Improvement of ant colony algorithm in group teaching and its application in robot path planning[J].CAAI transactions on intelligent systems,2022,17(4):764-771. Improvement of ant colony algorithm in group teaching and its application in robot path planning PU Xingcheng2,SONG Xinlin' (1.School of Computer Science and Technology,Chongqing University of Posts and Telecommunications,Chongqing 400065, China;2.School of Science,Chongging University of Posts and Telecommunications,Chongqing 400065,China) Abstract:To solve the problems of slow convergence speed and easily falling into local optimization,a novel improved ant colony algorithm is proposed based on a group teaching optimal algorithm(GTACO).The improved ant colony al- gorithm is optimized in three aspects.Firstly,the group teaching optimization algorithm is used to improve the fitness function of the ant colony algorithm to enhance the searching ability of global solutions.Simultaneously,a new fallback strategy is designed to deal with the U-shaped deadlock and ensure the feasibility of the solution.Secondly,a novel up- dating strategy for dynamic pheromones is adopted to avoid falling into local optimization of the algorithm by updating the path pheromone value after each iteration.Finally,the simplification operator of the path is applied to shorten the length of the path by simplifying the redundant corners into linear paths.Simulation experiments show that the im- proved algorithm can effectively increase the ability of path planning in convergence speed and accuracy for mobile ro- bots. Keywords:improved ant colony algorithm;group teaching optimization;path planning;mobile robot;pheromone up- date:heuristic function;path simplification;regression strategy 路径规划是机器人导航基础技术之一。传 收稿日期：2021-08-17.网络出版日期：2022-05-18. 基金项目：国家自然科学基金项目(61876200)：重庆市科委项 统路径规划算法有Dijkstra算法，A*算法等， 目(cstc2018 jeyjyAX0112):重庆市教委科研项目 (J2014032). 这些算法是基于图搜索路径规划算法。随着算法 通信作者：蒲兴成.E-mail:puxc@cqupt..edu.cnm 理论发展，基于智能优化路径规划算法被广泛应DOI: 10.11992/tis.202108020 网络出版地址: https://kns.cnki.net/kcms/detail/23.1538.TP.20220518.0930.002.html 分组教学蚁群算法改进及其在机器人路径规划中应用 蒲兴成1,2，宋欣琳1 （1. 重庆邮电大学 计算机科学与技术学院，重庆 400065; 2. 重庆邮电大学 理学院，重庆 400065） 摘 要：针对蚁群算法收敛速度慢、易陷入局部最优问题，提出一种基于分组教学优化改进蚁群算法。该算法 从 3 个角度对蚁群算法进行改进。首先，利用分组教学优化算法改进蚁群算法适应度函数，提高算法全局求解 能力。同时，引进一种新的回退策略，通过该策略处理 U 型障碍死锁问题，确保算法求解可行性。其次，采用 一种新的动态信息素更新策略，滚动更新每轮迭代后路径信息素值，避免算法陷入局部最优。最后，引入路径 简化算子，将冗余角简化为直线路径，缩短路径长度。仿真实验证明改进算法能有效提高移动机器人路径规划 收敛速度和精度。 关键词：改进蚁群算法；分组教学优化；路径规划；移动机器人；信息素更新；启发式函数；路径简化；回退策略 中图分类号：TP273 文献标志码：A 文章编号：1673−4785(2022)04−0764−08 中文引用格式：蒲兴成, 宋欣琳. 分组教学蚁群算法改进及其在机器人路径规划中应用 [J]. 智能系统学报, 2022, 17(4): 764–771. 英文引用格式：PU Xingcheng, SONG Xinlin. Improvement of ant colony algorithm in group teaching and its application in robot path planning[J]. CAAI transactions on intelligent systems, 2022, 17(4): 764–771. Improvement of ant colony algorithm in group teaching and its application in robot path planning PU Xingcheng1,2 ，SONG Xinlin1 (1. School of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, China; 2. School of Science, Chongqing University of Posts and Telecommunications, Chongqing 400065, China) Abstract: To solve the problems of slow convergence speed and easily falling into local optimization, a novel improved ant colony algorithm is proposed based on a group teaching optimal algorithm (GTACO). The improved ant colony al￾gorithm is optimized in three aspects. Firstly, the group teaching optimization algorithm is used to improve the fitness function of the ant colony algorithm to enhance the searching ability of global solutions. Simultaneously, a new fallback strategy is designed to deal with the U-shaped deadlock and ensure the feasibility of the solution. Secondly, a novel up￾dating strategy for dynamic pheromones is adopted to avoid falling into local optimization of the algorithm by updating the path pheromone value after each iteration. Finally, the simplification operator of the path is applied to shorten the length of the path by simplifying the redundant corners into linear paths. Simulation experiments show that the im￾proved algorithm can effectively increase the ability of path planning in convergence speed and accuracy for mobile ro￾bots. Keywords: improved ant colony algorithm; group teaching optimization; path planning; mobile robot; pheromone up￾date; heuristic function; path simplification; regression strategy 路径规划是机器人导航基础技术之一[1-4]。传 统路径规划算法有 Dijkstra 算法[5] ，A*算法[6] 等， 这些算法是基于图搜索路径规划算法。随着算法 理论发展，基于智能优化路径规划算法被广泛应 收稿日期：2021−08−17. 网络出版日期：2022−05−18. 基金项目：国家自然科学基金项目（61876200）；重庆市科委项 目 （cstc2018jcyjyAX0112）；重庆市教委科研项目 （J2014032）. 通信作者：蒲兴成. E-mail: puxc@cqupt.edu.cn. 第 17 卷第 4 期 智 能 系 统 学 报 Vol.17 No.4 2022 年 7 月 CAAI Transactions on Intelligent Systems Jul. 2022