一种改进的一维搜索指数优化算法  

An Improved Exponential Optimization Algorithm of One-Dimensional Search

作　　者：张志彬[1] 金福江[1] 汤仪平[2,3] 

机构地区：[1]华侨大学信息科学与工程学院,福建厦门361021 [2]华侨大学机电及自动化学院,福建厦门361021 [3]福建凤竹纺织科技股份有限公司,福建泉州362200

出　　处：《华侨大学学报（自然科学版）》2012年第5期503-505,共3页Journal of Huaqiao University(Natural Science)

基　　金：福建省产学研重大科研基金资助项目(2011H6019)

摘　　要：在分析黄金分割法基本原理的基础上,通过改变以指数收敛的区间长度缩短比率得到一种新的一维搜索指数优化算法.实例结果表明:该算法的收敛速度要比黄金分割法的收敛速度要快,同时最优解的区间精度也比黄金分割法的要精确;然而,该算法只适用于单峰函数局部最优解的求取.Based on the analysis of the basic principle of the golden section method,a new method called exponential optimization algorithm for one-dimensional search was presented by changing the interval length ratio of the exponential convergence.The results show that the method has faster convergence rate than that of the golden section method and the precision interval of the optimal solutions are also better than that of the golden section method.However,this algorithm applies only to calculate the local optimal solution of one-humped function.

关 键 词：一维搜索 黄金分割法 加速收敛 指数优化算法 

分 类 号：O232[理学—运筹学与控制论]