摘要
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,其中各宇宙独立演化．宇宙内采用免疫量子进化算法,宇宙间采用基于学习机制的移民、模拟量子纠缠的种群交叉等信息交互方式．使得进化算法具有更好的种群多样性,更快的收敛速度和全局寻优能力。仿真实验结果表明该算法比串行的免疫量子进化算法运算效率更高。
A novel multi-universe parallel immune quantum evolution algorithm（MPMQEA） based on the learning mechanism is proposed. In the algorithm, all individuals are divided into some independent subcolonies, called the universes. The topological structure of universes is defined. Each universe evolves independently. And the immune quantum evolution algorithm（MQEA） is applied to each of them. Information among universes is exchanged by adopting the emigration based on the learning mechanism and the quantum-cross simulating entanglement of the quantum. It can maintain the better population diversity, and help to accelerate the convergence speed and converge to the global optimal solution rapidly. Simulation results show that the efficiency of MPMQEA is higher than that of the serial MQEA.
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2006年第B07期79-82,共4页
Journal of Nanjing University of Aeronautics & Astronautics
基金
国家自然科学基金(60575040)资助项目。
关键词
并行量子进化算法
免疫量子进化算法
进化算法
量子计算
parallel quantum evolution algorithm
immune quantum evolution algorithm
evolution algorithm
quantum computing
