基于自适应变异粒子群优化算法的移动机器人路径规划  

Path Planning for Mobile Robot Based on Adaptive Mutation Particle Swarm Algorithm

作　　者：陈世明[1] 江冀海[1] 郑丽楠[1] 聂森[1] 

机构地区：[1]华东交通大学电气学院,江西南昌330013

出　　处：《机床与液压》2010年第23期16-19,共4页Machine Tool & Hydraulics

基　　金：国家自然科学基金项目(60804066);江西省自然科学基金项目(2007GQS0885);江西省教育厅科技项目(GJJ09508)

摘　　要：对路径规划中区域划分的建模方法进行改进,实现根据环境中障碍物的具体分布来自适应划分环境空间,将路径规划问题转化为带约束目标函数优化求解问题。针对粒子群算法的局部极小值问题,提出将自适应差分进化算子作为粒子群优化算法的自适应变异算子的改进方案。仿真结果表明:环境模型的自适应划分可以提高算法的搜索效率,缩短计算时间,使算法具有更好的实时性;加自适应变异算子的粒子群优化算法可以使目标函数快速跳出局部最优的束缚。A region division modeling method was improved for path planning of mobile robot,where adaptive division was realized according to the obstacles distribution in the environment.In this way,path planning problem was changed to be a constraint function optimization problem.To solve the problem of local optimum introduced by particle swarm optimization（PSO）algorithm,one modified scheme was proposed by taking adaptive differential evolution operator as adaptive mutation operator of particle swarm algorithm.The simulation results show：adaptive division of the environment model can improve searching efficiency,shorten calculating time,and make the algorithm possess better performance;the PSO adding adaptive mutation operator can quickly make the objective function out of the local optimum.

关 键 词：路径规划 局部极小值 自适应变异算子 粒子群算法 自适应差分进化算法 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]