摘要
为了减少机器人导航路径长度和优化时间,提出了基于蛙跳多种群粒子群算法的路径规划方法。建立了机器人工作环境的Maklink模型,首先使用MS算法搜索出若干最短路径,然后提出了蛙跳多种群粒子群算法进行路径二次优化。将蛙跳算法的深度搜索思想引入到粒子群算法中,提出了多种群粒子群算法的分群方法、更新策略和合作机制,进而给出了基于蛙跳多种群粒子群算法的机器人导航路径优化方法。经过仿真验证,蛙跳多种群粒子群算法具有最佳的优化效果,最短路径长度比MSCPSO算法减少了3.82%,比PSO算法减少了5.46%;另外,蛙跳多种群粒子群算法的运行时间比MSCPSO算法减少了25.53%,比PSO算法减少了18.79%。
In order to reduce robot working path length and algorithm running time,path planning method based on leapfrog multi-group PSO is proposed. Maklink model of robot working environment is built,and several shortest path are searched by“MS algorithm in essay,then the secondary optimization is executed by leapfrog multi-group PSO.Detailed search thought in shuffled leapfrog algorithm is introduced to PSO,and group-dividing method,update strategy and cooperative mechanism of multigroup PSO is put forward. So that robot path planning method based on leapfrog multi-group PSO is put forward. Clarified by simulation,leapfrog multi-group PSO possess optimal searching property. Length of shortest path by LMGPSO decreases by3.82% compared with MSCPSO,and 5.46% compared with PSO.Running time of LMGPSO decreases by 25.53% compared with MSCPSO,and 18.79% compared with PSO.
作者
李硕
苏鸣
赵燕
LI Shuo;SU Ming;ZHAO Yan(Wuchang Shouyi University,Hubei Wuhan 430064,China;Wuhan University of Science and Technology,Hubei Wuhan 430080,China)
出处
《机械设计与制造》
北大核心
2022年第3期258-261,265,共5页
Machinery Design & Manufacture
基金
湖北省教育厅科技指导项目(B2017364)。
