计算机与现代化 ›› 2011, Vol. 1 ›› Issue (8): 146-148.doi: 10.3969/j.issn.1006-2475.2011.08.040

• 计算机仿真 • 上一篇    下一篇

关于计算机配色模型参数值的优化方法讨论

韩 蔚,丁慎军   

  1. 莱芜职业技术学院计算机系,山东 莱芜 271100
  • 收稿日期:2011-03-28 修回日期:1900-01-01 出版日期:2011-08-10 发布日期:2011-08-10

Discussion of Optimization Methods About Computer Matching Model Parameters

HAN Wei, DING Shen-jun   

  1. Computer Department, Laiwu Vocational and Technical College, Laiwu 271100, China
  • Received:2011-03-28 Revised:1900-01-01 Online:2011-08-10 Published:2011-08-10

摘要: 就数学模型中参数值的优化问题列出两种不同的优化方法——非线性规划函数寻优和遗传算法最优化工具箱中的gaopt函数寻优。针对两种优化方法各自的特点,把gaopt函数的最优解作为非线性规划函数寻优的初始值,解决非线性规划函数初始值难以确定的问题,同时也能更快速、更合理地找到所需的参数值。结果表明,采取将遗传算法同函数调用相结合的方法,可较好地实现模型中参数值的优化。

关键词: 函数调用, 遗传算法, 最优化, 配色系数, 非线性

Abstract: This article offers two different optimization methods as for the problem of parameter’s optimization in the mathematics model. That is, non-linear programming function and the function of gaopt in the genetic algorithm optimization toolbox. In view of respective characters of the two different optimization methods, the optimal value of gaopt function is regarded as the initial value of optimization of non-linear programming functions. The method solves the problem about the uncertainty of the initial value of non-linear programming functions and the required parameter values would be quickly and reasonably searched. The result proves that the method of the combination of genetic algorithm and function transfer preferably realizes the optimization of parameter values in mathematics model.

Key words: function transfer, genetic algorithm, optimization, matching coefficient, non-linear

中图分类号: