函数优化问题的动态并行量子遗传算法  

Dynamic Parallel Quantum Genetic Algorithm for Function Optimization Problem

作　　者：马胡双 石永革[1] 

机构地区：[1]南昌大学信息工程学院,江西南昌330031

出　　处：《青岛科技大学学报（自然科学版）》2017年第1期109-115,共7页Journal of Qingdao University of Science and Technology:Natural Science Edition

基　　金：国家自然科学基金项目(61163005)

摘　　要：针对传统量子遗传算法在解复杂连续函数优化中存在的早熟收敛、收敛速度慢、计算时间长的问题,提出一种解复杂连续函数优化问题的动态并行量子遗传算法(DPQGA)。采用多种群协同进化,每个子种群按照各自的进化目标在不同的搜索区域进化,形成并行搜索方式,加快算法收敛速度,避免早熟收敛;同时设计了一种新的动态量子旋转角的更新策略及量子门调整策略,减少算法的迭代次数;在最优解连续数代无变化时引入灾变算子,使种群保持良好的多样性。通过对5个测试函数的仿真,结果表明,该算法搜索到的最优解较QGA算法更优。与已有算法相比,该算法在收敛速度、迭代次数、全局寻优能力上都有了较大的改进和提高。For that traditional quantum genetic algorithm exist the problem of premature convergence,slow convergence speed and long computing time for complex continuous function optimization problem,a new algorithm named dynamic parallel quantum genetic algorithm（DPQGA）for complex continuous function optimization problem is proposed.Multiple population co-evolution is adopted.Each child population evolves in different search area according to the respective target.That will form parallel search way.So it can accelerate the algorithm convergence speed and avoid premature convergence.Meanwhile,a kind of new dynamic update strategy of quantum rotation Angle and quantum gate adjustment strategy is designed.It can reduce the number of iterations of the algorithm.When there is no change in the optimal solution for several generations,cataclysm operator is introduced into proposed algorithm to keep good population diversity.Through the experiment of five test functions,the results show that the optimal solutions of proposed algorithm are more optimal thanQGA.Compared with existing algorithms,the convergence speed,the number of iterations and the global optimization ability of the algorithm has a bigger improvement and improve.

关 键 词：复杂连续函数优化 量子遗传算法 动态调整旋转角 协同进化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]