摘要
有别于传统的单目标方法,将饲料配方设计问题描述成为一个多目标最优化问题,并提出了一种改进的多目标遗传算法,应用模拟退火算法解决多目标遗传算法的局部搜索能力差和易早熟问题.实验结果表明,该算法能有效地求解饲料配方设计问题并为设计者提供了强有力的决策支持.
Unlike traditional single objective method, feed formulation design is represented as a multi-objective optimization problem. An improved multi-objective genetic algorithm for feed formulaiton design is presented. Simulated annealing algorithm is presented to overcome the problems of multi-objective genetic algorithm, such as poor local search ability and premature convergence. The results indicate that this algorithm can effectively solve the problems in feed formulation design and provide powerful decision support for the designer.
出处
《河南工业大学学报（自然科学版）》
CAS
北大核心
2009年第3期75-78,共4页
Journal of Henan University of Technology：Natural Science Edition
关键词
遗传算法
多目标优化
模拟退火
饲料配方设计
genetic algorithm
multi-objective optimization
simulated annealing
feed formulation design
