| 一类非线性约束优化的修正正割方法的收敛性 |
| |
| 引用本文： | 朱德通.一类非线性约束优化的修正正割方法的收敛性[J].高校应用数学学报(A辑),1992,7(4):531-543. |
| |
| 作者姓名： | 朱德通 |
| |
| 作者单位： | 上海师范大学数学系 邮码 |
| |
| 基金项目： | 上海师范大学自然科学基金 |
| |
| 摘 要： | 本文修正了一类非线性约束优化的正割方法.通过引入不可微势函数作不精确的一维搜索,证明了修正后的正割算法不仅具有原算法不具备的整体收敛性;而且保持局部两步Q-超线性收敛速率.进一步数值结果表明此算法是非常有效的.
|
| 关 键 词： | 约束优化 势函数 非线性 正割法 |
CONVERGENCE OF A CLASS OF IMPROVED SECANT METHODS FOR NONLINEAR CONSTRAINED OPTIMIZATION |
| |
| Zhu Detong.CONVERGENCE OF A CLASS OF IMPROVED SECANT METHODS FOR NONLINEAR CONSTRAINED OPTIMIZATION[J].Applied Mathematics A Journal of Chinese Universities,1992,7(4):531-543. |
| |
| Authors: | Zhu Detong |
| |
| Institution: | Shanghai Normal University |
| |
| Abstract: | In this paper we improve a class of secant methods for nonlinear constrained optimization in 1], by introducing a nonsmooth merit function to make inexact one-dimensional searches. It is shown that improved secant methods have gained global convergence property which is not possessed by the original secant methods, and it is also proved that they are locally two-step Q-superlinarly convergent. Furthermore, numerical tests show that the improved secant methods are very effective. |
| |
| Keywords: | Secant Methods Constrained Optimization Merit Function Superlinear Convergence Global Convergence |
| 本文献已被 CNKI 维普 等数据库收录！ |
