| 摘要: |
| 利用广义投影技术和次可行方向法思想建立了非线性等式与不等式约束最优化问题的一个算法.它采用广义投影代替了传统的转轴运算,而且广义投影阵只由ε-积极约束函数的梯度产生.对于不等式约束的辅助优化问题,该算法是一个次可行方向类算法,称之为广义次可行方向法.算法在较弱的条件下具有全局收敛性. |
| 关键词: 非线性等式与不等式约束 最优化问题 广义投影 广义次可行方向法 全局收敛性 |
| DOI： |
| 投稿时间：1995-09-20 |
| 基金项目:广西自治区青年科学基金和广西区教委资助项目。 |
|
| A Generalized Subfeasible Directions Method for Optimization Problems Using Generalized Projection |
|
Jian Jinbao
|
| (Dept. of Math. & Info. Sci., Guangxi Univ., 10 Xixiangtanglu, Nanning, Guangxi, 530004) |
| Abstract: |
| An algorithm for optimization problems with nonlinear equality and inequality constraints is presented by using the generalized projection and subfeasible directions method.It uses the generalized projection to replace the traditional pivotal operations,and the generalized projection matrices can be generated only by the gradients of constrained functions within the ε-active constrained set.This method is a subfeasible directions method for the inequality constraints assistant problem, we call it generalized subfeasible directions method.The algrithm possesses global convergence under weaker condictions. |
| Key words: nonlinear equality and inequality constraints optimization problems generalized projection generalized subfeasible directions method global convergence |
