机构地区: 广东石油化工学院
出 处: 《计算机测量与控制》 2011年第8期2022-2025,共4页
摘 要: 为解决传统遗传算法收敛速度慢、群体多样性不足的缺陷,提出了一种多策略并行的遗传算法;算法采用多策略并行处理的方式,产生不同策略模式下的个体,增加群体的多样性,再经过自适应迁移策略,提高算法的收敛速度;利用Markov链模型分析多策略并行遗传算法的收敛性;采用旅行商函数进行算法测试,结果表明改进算法的收敛性较传统遗传算法有较大的提高,具有较强的工程应用性能。 To deal with slower convergence speed and the deficiency of population diversity of traditional genetic algorithm, a multi-- strategy parallel algorithm is proposed. The parallel processing of multi--strategy is used to produce the individual under the different strategy modes, increase the population diversity, then the convergence speed is improved after adaptive migration strategy. The Markov chain model is used to analyze the convergence of multi--strategy parallel algorithm. TSP function is made use of algorithm testing, experimental results shows that compared with the traditional genetic algorithm, the improved algorithm raised the eonvergence, and it has a strong application performance.
领 域: [自动化与计算机技术] [自动化与计算机技术]