基于混沌粒子群——专用遗传算法切换策略的移动机器人路径规划  

Switch strategy based on chaos particle swarm optimization and specialized genetic algorithm for path planning of mobile robots

作　　者：张超[1] 李擎[1] 董冀媛[1] 韩彩卫[1] 刘启晗[1] 

机构地区：[1]北京科技大学自动化学院,北京100083

出　　处：《北京科技大学学报》2013年第6期826-830,共5页Journal of University of Science and Technology Beijing

基　　金：教育部第36批留学回国人员科研启动基金资助项目(1341);国家自然科学基金资助项目(60374032);北京市重点学科建设资助项目(XK100080537)

摘　　要：为了发挥粒子群算法和专用遗传算法的各自优点,提出了一种将二者结合的切换优化策略.该策略前期采用一种基于种群最优个体混沌化的混沌粒子群算法,后期选用专用遗传算法.通过大量仿真实验确定了在迭代代数、种群标准差和最优个体适应度差三种切换指标下各自的最优切换条件.与单一专用遗传算法和单一混沌粒子群算法的仿真对比表明:本文提出的切换优化策略在综合路径长度、平滑性和规划时间三个性能指标后具有一定的优越性.A switching strategy based on （CPSO-SGA） was presented by combining their chaos particle swarm optimization own advantages. In the switching and specialized genetic algorithm strategy, CPSO is applied in the former step and SGA is executed in the later step. The best switching conditions under three switching indices of iteration steps, population standard deviation, and optimal individual fitness values were determined by large amounts of simulation experiments. In comparison with single SGA and single CPS0, the proposed switching strategy CPSO-SGA has a better performance when path length, smoothness, and running time are taken into consideration.

关 键 词：移动机器人 路径规划 粒子群算法 遗传算法 切换 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]