文献类型：期刊文献

中文题名：并行量子进化算法的研究与实现

英文题名：RESEARCH AND REALIZATION OF PARALLEL QUANTUM EVOLUTIONARY ALGORITHM

作者：游晓明[1];刘升[1];帅典勋[2]

机构：[1]上海工程技术大学电子电气工程学院,上海200065;[2]华东理工大学计算机科学与技术系,上海200237

年份：2008

卷号：25

期号：5

起止页码：231

中文期刊名：计算机应用与软件

外文期刊名：Computer Applications and Software

收录：CSTPCD;;北大核心:【北大核心2004】；CSCD:【CSCD_E2011_2012】；

语种：中文

中文关键词：并行进化模型;量子进化算法;免疫算子;交叉变异

外文关键词：Parallel evolutionary model Quantum evolutionary algorithm Immune operator Cross-mutation

摘要：分析讨论并行进化模型理论及性能,提出了基于学习的多宇宙并行免疫量子进化算法,算法中将种群分成若干个独立的子群体,称为宇宙。并给出了多宇宙的并行拓扑结构,提出了宇宙内采用免疫量子进化算法,宇宙之间采用基于学习的移民和模拟量子纠缠的交互策略进行信息交换。这样能提高种群多样性,有效克服早熟收敛现象。算法综合了量子计算的天然并行性和免疫算法的充分自适应性,它比传统的进化算法具有更好的种群多样性,更快的收敛速度。通过并行实验验证了该算法的优越性。
The paper analyzes and discusses the theory of parallel evolutionary model and its performance, and educes Multi-universe Parallel Immune Quantum Evolutionary Algorithm （MPMQEA） based on learuing mechanism. In the algorithm, population was divided into a number of independent sub-populations which were called universes, the topological structure of multi-universe was defined： It is proposed to use immune quantum evolutionary algorithm inside each univeise while to use the learning mechanism based emigration and the interaction policy of simulating quantum entanglement for quantum information exchange among the universes. In this way the population diversity is enhanced, the premature convergence is overcome effectively. The intrinsic parallelism of quantum evolutionary algorithm was integrated with full adaptivity of immune dynamic model in MPMQEA to make it have higher population diversity and quicker convergence speed than traditional evolutionary algorithm. Its superiorities were verified by the simulation experiments.