一种新的饲料配方技术——杂种遗传算法

A New Procedure Used for Feed Formulation: Hybrid Genetic Algorithm

本文旨在研究杂种遗传算法应用于饲料配方设计的可行性。利用杂种遗传算法与已经应用的6种饲料配方算法(标准线性规划、目标规划、模糊线性规划、随机规划、模拟退火搜索和标准遗传算法)比较,结果显示,杂种遗传算法的配方满足了营养需要量和原料成分的约束要求,配方成本也有一定程度的减少。实例表明,杂种遗传算法应用于饲料配方设计具有可行性。

This study was conducted to study the feasibility of hybrid genetic algorithm apply to the design of feed formulation. Hybrid genetic algorithm （HGA） was compared with 6 feed formulation methods in using： general linear programming （LP）, goal programming （GP）, fuzzy linear programming （FLP）, stochastic programming （SP）, simulated annealing research algorithm （SAA）, and general genetic algorithm （GA）. The results showed that hybrid genetic algorithm was satisfied with the constrained nutrient requirements and the ingredient restrictions, and costs of formula ingredients were reduced by different sizes also. The illustrations showed that hybrid genetic algorithm had the feasibility in the feed formulation. [Chinese Journal of Animal Nutrition, 2009,21 （5） ：703-710 ]