基于改进遗传算法的饲料配方多目标优化研究
Multi-objective formulation optimization study based on improved genetic algorithms
龙艳1
作者信息
- 1. 新疆应用职业技术学院,新疆 奎屯 833200
- 折叠
摘要
为提高动物饲料配方的营养,降低饲料成本,在保证饲料营养的前提下,构建最小成本的目标函数,并采用改进遗传算法对目标函数的最优配方进行求解.仿真实验表明,以 20～50 kg的生猪为实验范例,分别采用标准遗传算法和改进遗传算法求解 100 kg生猪饲料配方,并将两种方法求解得到的最终饲料配方进行对比.实验结果表明:通过标准遗传算法求解获得的 100 kg生猪饲料配方的总成本为 307.91 元,而改进的遗传算法求解获取的 100 kg生猪饲料配方的总成本为 291.03 元,比传统遗传算法获取的配方减少了 16.88 元;标准遗传算法在经过 650 次迭代后陷入了局部最优解,而改进遗传算法在经过 700 次迭代后变化曲线趋于平稳,输出全局最优解291.03 元.由此得出,改进遗传算法的全局寻优能力更强,可快速、有效求解动物饲料最优配方.
Abstract
To improve the nutrition of animal feed formulas and reduce feed costs,a minimum cost objective function is con-structed while ensuring feed nutrition,and an improved genetic algorithm is used to solve the optimal formula of the objective function.The simulation experiment shows that using 20-50 kg live pigs as experimental examples,the standard genetic algo-rithm and improved genetic algorithm are used to solve the 100 kg pig feed formula,and the final feed formula obtained by the two methods is compared.The experimental results show that the total cost of the 100 kg pig feed formula obtained through standard genetic algorithm is 307.91 yuan,while the total cost of the 100 kg pig feed formula obtained through improved genetic algorithm is 291.03 yuan,which is 16.88 yuan less than the formula obtained through traditional genetic algorithm;The stand-ard genetic algorithm fell into a local optimal solution after 650 iterations,while the improved genetic algorithm stabilized its change curve after 700 iterations,outputting a global optimal solution of 291.03 yuan.From this,it can be concluded that the improved genetic algorithm has stronger global optimization ability and can quickly and effectively solve the optimal formula for animal feed.
关键词
机器学习/遗传算法/饲料配方/饲料成本Key words
machine learning/genetic algorithms/feed formulation/feed cost引用本文复制引用
出版年
2024
NETL
NSTL
万方数据