針對(duì)遺傳算法局部搜索能力弱的缺陷，提出了一種改進(jìn)的混合遺傳算法。根據(jù)遺傳的不同階段分為兩個(gè)不同的群體——競(jìng)爭(zhēng)群體和適應(yīng)性群體，提出相關(guān)的遺傳算子——繁殖因子。將運(yùn)籌學(xué)中的單純形法應(yīng)用于遺傳算法中，增強(qiáng)了遺傳算法的局部搜索能力。對(duì)復(fù)雜函數(shù)的尋優(yōu)實(shí)驗(yàn)驗(yàn)證了混合遺傳算法的有效性，并通過(guò)與傳統(tǒng)SGA的實(shí)算結(jié)果對(duì)比，更進(jìn)一步說(shuō)明了算法的改進(jìn)效果。There are some limitations that using genetic algorithms. This paper presents the Mixed Genetic Algorithm(MGA). Based on the phase of genetic，the colonies are differentiated to two different colonies-adaptive population and competing population. The correlated operator, propagate gene, is brought forward. The local search ability of the genetic algorithm is enhenced by using the simplex algorithm of operational research. The emulation experiment data shows the optimizing convergence reliability and higher converging velocity.