求解无约束全局优化的改进的单填充函数法  

A Modified Single-Parameter Filled Function Method for Unconstrained Global Optimization

作　　者：骆世云[1] 叶仲泉[1] 

机构地区：[1]重庆大学数理学院,重庆400030

出　　处：《计算机技术与发展》2008年第8期108-110,151,共4页Computer Technology and Development

摘　　要：填充函数法是一种求解多变量、多极值函数全局最优化的有效方法,这种方法的关键是构造填充函数。为此文中根据文献[1]的思想,考虑优化问题minf(x)The filled function method is an effective approach for finding the global minima of multirnodal and multidimensional functions, and the constructed filled function is vital to the results of optimization. Therefor, by the thought of the literature[1], considering the optimization problem min f（x）x∈R^n, when f（ x ） is local Lipschitz continuous function, proposed a modified single - parameter filled function. It is easy to prove that this new filled function can keep filling properties easily relative to conventional filled functiuns , furthermore, its global convergent speed is rapid. The corresponding algorithm was also discussed in this paper. Numerical experience for 4 test functions indicate that the new method is better than the method in the reference.

关 键 词：填充函数法 全局优化 数值实验 

分 类 号：O221.2[理学—运筹学与控制论]