文献类型：期刊文献

中文题名：求解箱式约束全局优化问题的降维算法

英文题名：Dimensionality Reduction Algorithm for Global Optimization Problems with Closed Box Constraints

作者：王薇[1];袁琪[1];唐加会[1]

机构：[1]华东理工大学理学院,上海200237

年份：2019

卷号：8

期号：1

起止页码：38

中文期刊名：数学建模及其应用

外文期刊名：Mathematical Modeling and Its Applications

基金：国家自然科学基金(71531005)

语种：中文

中文关键词：全局最优化;降维变换;α-致密;加速单变量算法

外文关键词：global optimization problems;dimensionality reduction;α-dense;accelerated single variable method

摘要：本文利用有限区间降维方法,将带箱式约束的多维优化问题转化为一维优化问题.然后利用一种加速方法对一维优化问题求全局最优解,并证明该最优解是原问题的近似解.最后给出算法和数值算例结果.
In this paper,the multi-dimensional optimization problem with box constraints is transformed into onedimensional optimization problem by using the finite interval dimension reduction method.The global optimal solution of the one-dimensional optimization problem is obtained by an accelerated method,which is an approximate solution of the original problem.Also the algorithm and the numerical results are listed.