| 徐莹莹,张惠.一类改进的拟牛顿算法[J].井冈山大学自然版,2018,(1):21-23 |
| 一类改进的拟牛顿算法 |
| A CLASS OF NEW MODIFIED QUASI-NEWTON ALGORITHM |
| 投稿时间：2017-04-11 修订日期：2017-09-20 |
| DOI：10.3969/j.issn.1674-8085.2018.01.005 |
| 中文关键词: 无约束优化 拟牛顿方程 线性搜索准则 全局收敛性 |
| 英文关键词: unconstrained optimization quasi-Newton equation linear search global convergence |
| 基金项目: |
|
| 摘要点击次数: 10001 |
| 全文下载次数: 18823 |
| 中文摘要: |
| 在利用拟牛顿算法求解非线性无约束优化问题中，本文在文献[8]提出的拟牛顿方程基础上，通过加权形式构造一类改进拟牛顿方程，产生了修正的BFGS校正公式，进而提出改进的拟牛顿算法，在一定条件下证明新算法的全局收敛性。数值实验结果表明，与文献[12]中的拟牛顿算法对比，新算法在迭代次数上更有优势。 |
| 英文摘要: |
| In solving nonlinear unconstrained optimization problem by using quasi-Newton algorithm. Though the weighted form a class of new quasi-newton algorithm is constructed based on the quasi-newton equation and the method of literature. Furthermore, a new quasi-newton algorithm is proposed combining the modified BFGS correction formula of the new quasi-newton equation. The new global convergence of the algorithm is proved under certain conditions. Finally, through numerical experiments show that this new algorithm has much more advantages in the number of iterations. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
