摘 要:
将一般的无约束非单调线搜索算法推广成凸约束的非单调线搜索法,搜索方向由BFGS方法改变为满足凸约束条件的方程组来确定,证明了算法的全局收敛性和k次线性收敛,进行了数值试验,结果表明算法十分有效。[著者文摘]
文章出处:
《常州工学院学报》-2007年20卷5期,64 -52-56,64页
Journal of Changzhou Institute of Technology
栏目信息:
分 类 号:
文献标识码:
A
文章编号:
1671-0436(2007)02-0052-05
相关文章:
A Nonmonotone Line Search Method for Convex Constrained Optinization
XIA Hong-wei ( School of Science, Changzhou Institute of Technology, Changzhou 213002)
Abstract:
An unconstrained nonmonotone line search algorithm is generalized to the one with convex constrained, the search direction dk determined by BFGS method is replaced by the one determined by the equations which satisfy the convex constrained conditions, and testify the global convergence and k-linear convergence for this algorithm. The numerical test shows the new method is very effective.[著者文摘]
Key words:
convex constrained; line search; convergence.
收稿日期: 2007-09-07

更多评论>>文章评论
学术















cqvip.com