复杂多陷阱环境下机器人导航路径的蚁群规划策略  

Robot Navigation Path Planning Strategy Under Complex Multiple Traps Environment by Ant Colony Algorithm

作　　者：李婧[1] 李艳萍[2] LI Jing;LI Yan-ping(Shanxi Polytechnic College,Shanxi Taiyuan 030006,China;School of Information and Computer Engineering,Taiyuan University of Technology,Shanxi Taiyuan 030024,China)

机构地区：[1]山西职业技术学院,山西太原030006 [2]太原理工大学信息与计算机工程学院,山西太原030024

出　　处：《机械设计与制造》2023年第8期228-232,共5页Machinery Design & Manufacture

基　　金：山西省回国留学人员科研资助项目(2017-031)。

摘　　要：为了提高复杂多陷阱环境下的机器人路径规划质量和规划效率,提出了多因素启发蚁群算法的路径规划方法。使用栅格法建立了工作环境的“0-1”矩阵模型,同时给出了环境模型的转移图结构。分析了蚂蚁进入陷阱时“蚂蚁回退策略”和“蚂蚁夭折策略”的缺陷,针对这些缺陷,将安全启发因素和智能蚂蚁策略融入到蚁群算法中,提出了能够高效应对陷阱问题的多因素启发蚁群算法,并将其应用于复杂多陷阱环境下的路径规划。在3种仿真环境下对改进措施进行验证,由实验结果可以看出,多因素引导蚁群算法能够有效避开障碍物密集区域;在复杂多陷阱环境下改进算法规划的路径长度明显短于传统算法,路径拐点远少于传统算法,迭代次数明显少于传统算法。以上实验结果表明,多因素启发蚁群算法的规划质量和规划效率均优于传统蚁群算法。In order to improve robot path planning quantity and efficiency under complex multiple traps environment,path planning method based on multiplefactors guidance ant colony algorithm is proposed.O-1 matrix model ofenvironment is built by grid method,and transferring figure structure of environment model is given at the same time.Shortcomings of ant backspacing strategy and ant mortality strategy are analyzed,direct at the shortcomings,safety inspiration factor and intelligent ant strategy are introduced to ant colony,so that multiple factors guidance ant colony algorithm which can reply traps effectively is put forward.The improved algorithm is used to path planning under multiple traps environment.Improved measures are clarified under3 simulation environments,the result shows that the improved algorithm can elude barrier-dense region.Under complex multiple traps environment,compared with traditional ant colony algorithm,path length planned by improved algorithm is shorter,quantity of path inflection point is less,and iteration time is obvious less than traditional algorithm.The experiment result indicates that path quality and efficiency of multiple factors guidance ant colony algorithm is optimal than traditional ant colony algorithm.

关 键 词：多陷阱环境 导航路径规划 安全引导因素 智能蚂蚁策略 多因素引导蚁群算法 

分 类 号：TH16[机械工程—机械制造及自动化] TP242[自动化与计算机技术—检测技术与自动化装置]