改进蚁群算法的群体机器人多目标搜索方法  

Improved Ant Colony Algorithm for Multi Objective Search of Group Robots

作　　者：刘艺美 张跃进[2] LIU Yimei;ZHANG Yuejin(Jiangxi Machinery Science Research Institute,Jiangxi Nanchang 330095,China;School of Information Engineering,East China Jiaotong University,Jiangxi Nanchang 330013,China)

机构地区：[1]江西省机械科学研究所,江西南昌330095 [2]华东交通大学信息工程学院,江西南昌330013

出　　处：《机械设计与制造》2024年第12期202-206,共5页Machinery Design & Manufacture

基　　金：国家自然科学基金重大研究计划培育项目(92159102);江西省教育厅科学技术研究项目(GJJ2200642)。

摘　　要：在多目标搜索中,群体机器人的目标搜索过程中,很容易出现相互重复搜索,主要以为依靠的蚁群算法在信息素更新过程中,存在重复操作,为此提出基于改进蚁群算法的群体机器人多目标搜索方法。针对蚁群算法中的信息素浓度更新规则,以最大最小交叉为原则改进,得到基于路径选择概率和路径长度的信息素更新方式。结合栅格化处理、目标孳息计算、搜索步长动态调整、状态转移概率与轮盘赌的最终单元选择等操作,完成群体机器人对多个目标的搜索任务。在静态与动态两种环境中展开测试,通过各机器人的间距、机器人与障碍物的间距、机器人与目标间到达情况,可知所提方法下群体机器人以理想的间距精准躲避了障碍物,且成功搜索到全部目标,方法应用效果较好。In multi-target search,it is easy for swarm robots to repeatedly search each other in the target search process,mainly because the Ant colony optimization algorithms relies on repeated operations in the pheromone update process,a multi-objective search method for swarm robots based on improved ant colony algorithm is proposed.For the pheromone concentration update rule in Ant colony optimization algorithms,the maximum and minimum cross principle is adopted to improve,and the pheromone update method based on path selection probability and path length is obtained.Combined with rasterisation processing,target fruits calculation,dynamic adjustment of search step size,state transition probability and final unit selection of roulette wheel gambling,the swarm robot can search multiple targets.Testing was conducted in both static and dynamic environments,through the distance between each robot,the distance between the robot and the obstacle,and the arrival situation between the robot and the target,it was found that the proposed method accurately avoided the obstacle at an ideal distance and successfully searched for all targets,the application effect of the method is good.

关 键 词：改进蚁群算法 信息素更新 群体机器人 轮盘赌策略 多目标搜索 

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]