一种改进的多种群量子遗传算法  

An improved multi-population quantum genetic algorithm

作　　者：王欢 WANG Huan(Network and Modern Education Technology Center, Zhongkai University of Agricuhure and Engineering, Guangzhou 510225, China)

机构地区：[1]仲恺农业工程学院网络与现代教育技术中心,广东广州510225

出　　处：《仲恺农业工程学院学报》2017年第4期40-45,共6页Journal of Zhongkai University of Agriculture and Engineering

摘　　要：针对常规量子遗传算法(Quantum genetic algorithm,QGA)在求解连续函数优化问题时容易陷入局部极值,提出了一种改进的多种群量子遗传算法(Improved multi-population quantum genetic algorithm,IMPQGA).该算法将初始化种群划分成N个子种群,每个子种群按不同的量子旋转门策略更新,然后相互交换子种群最优个体,同时在算法进化中引入一种新的量子旋转门,随进化代数增加动态地调整染色体个体进化方向,使算法及时跳出局部最优,避免早熟收敛.仿真结果表明,该算法相比常规量子遗传算法和多种群遗传算法(Multi-population quantum genetic algorithm,MPQGA)具有更好的优化性能.It was easy to fall into local optimum when using conventional quantum genetic algorithm to solved continuous function optimization. In order to surmount the above drawback, an improved multi-po- pulation quantum genetic algorithm （IMPQGA） was proposed. The algorithm divided the initial popula- tion into N sub population. Each population had updated itself respectively according to the different quantum rotation gate strategies, and then had exchanged the optimal individuals with the next sub popu- lation. At the same time, in order to jump out of local optimum in time and avoided premature conver- gence, a novel quantum rotation gate was introduced to dynamically adjust the chromosome evolution di- rection with the increasing number of evolution generation. The simulation results showed that IMPQGA had better optimization performance with the multi population quantum genetic algorithm （ MPQGA ）. Compared with the conventional quantum genetic algorithm （QGA）.

关 键 词：量子遗传算法 量子旋转门 改进型多种群量子遗传算法 连续函数优化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]