摘要
遗传算法是一种概率搜索算法,其本身固有并行性。目前,人们正不断地致力于把遗传算法应用于各种并行机器上。在基于工作站机群技术上,构架了一种粗粒度并行遗传算法,并在MPI并行环境下,用4台PC机测试了一个解决TSP问题的粗粒度并行遗传算法。该并行遗传算法可以更好的保护优秀个体从而提高遗传算法的收敛性。
Genetic Algorithm（GA）, with its innate parallelism, is a probability search algorithm. People are currently trying to apply GA to all kinds of parallel systems. Based on COW, a parallel genetic algorithm modal is designed. Under MPI it is tested on a COW cluster consisting of four PCs to deal with TSP. The Parallel Genetic Algorithm can protect the excellent individuals and therefore improve convergence.
出处
《北京联合大学学报》
CAS
2006年第2期40-43,共4页
Journal of Beijing Union University
