耿焕同, 冯慧菁, 贾婷婷. 一种多种群协同量子遗传多峰函数优化方法[J]. 微电子学与计算机, 2013, 30(11): 56-59,63.
引用本文: 耿焕同, 冯慧菁, 贾婷婷. 一种多种群协同量子遗传多峰函数优化方法[J]. 微电子学与计算机, 2013, 30(11): 56-59,63.
GENG Huan-tong, FENG Hui-jing, JIA Ting-ting. A Multi-Population Synergy Quantum Genetic Algorithm for Multimodal Optimization[J]. Microelectronics & Computer, 2013, 30(11): 56-59,63.
Citation: GENG Huan-tong, FENG Hui-jing, JIA Ting-ting. A Multi-Population Synergy Quantum Genetic Algorithm for Multimodal Optimization[J]. Microelectronics & Computer, 2013, 30(11): 56-59,63.

一种多种群协同量子遗传多峰函数优化方法

A Multi-Population Synergy Quantum Genetic Algorithm for Multimodal Optimization

  • 摘要: 量子遗传算法是一种基于概率的进化算法,在求解多峰函数优化问题时其旋转门的更新策略容易导致整个种群陷入局部最优,并且在有多个最优解的优化问题中不能找全最优解。针对以上不足,提出了一种基于多种群协同量子遗传算法(MPSGQA),来寻找多峰函数优化问题中的所有全局最优解。其思想是在初始种群中采用聚类分析方法划分 m个子种群,然后每个子种群分别进化更新,当子种群之间的相似性接近时采用量子变异策略保持种群的多样性。在多峰优化函数 F1~F4上进行实验与比较,结果表明,相比传统量子遗传算法,M PSGQA可以很好地跳出局部最优,并能找全多峰函数的多个最优解。

     

    Abstract: The quantum genetic algorithm is a kind of probability evolution algorithm but when it solves multimodal optimization problems,the update strategy of rotation gate will easy lead to local optimal solution and also can't find all best solutions in some optimization problem.Therefore in this paper a multi-population synergy quantum genetic algorithm (MPSGQA) proposed to solve these problems and find all the best solutions.The main idea is dividing m sub-populations by using clustering analysis and the update them. If some of the sub-populations have large similarity,the quantum mutation strategy is adopted to maintain the diversity of population. Experiment and comparison with multimodal optimization function F1 to F4,the result shows that MPSGQA can jump out of local optimal solution better than quantum genetic algorithm and find all global optimum solutions efficiently.

     

/

返回文章
返回