粗粒度并行遗传算法的计算性能分析  

Performance Analysis of the Coarse-grained Parallel Genetic Algorithms

作　　者：岳嵚[1] 冯珊[1] 

机构地区：[1]华中科技大学控制科学与工程系,武汉430074

出　　处：《武汉理工大学学报》2008年第7期107-110,共4页Journal of Wuhan University of Technology

基　　金：国家自然科学基金(60774084)

摘　　要：探讨了粗粒度并行遗传算法的机理和运行的基本步骤,通过对经典测试函数Bohachevsky 3#函数进行不同参数组合的多次计算并比较计算结果,分析了粗粒度并行遗传算法的计算性能特点。计算结果表明:粗粒度并行遗传算法相对于经典遗传算法有较为理想的运行结果和较为理想的运行过程。The paper discussed the mechanism and brief process of the coarse-grained parallel genetic algorithms （CPGA）. It computed the classical testing function, Bohachevsky 3# function many times with different parameter combinations. It analyzed the computing performance of CPGA by contrasting the computing results. The results of the computing indicated that CPGA not only achieve better performance than the canonical genetic algorithms in the computing result, but also achieve better performance in the computing process.

关 键 词：粗粒度并行遗传算法 子种群 种群多样性 计算稳定性 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]