摘要
在传统复合形直接搜索方法基础上,对有约束的非线性优化问题,充分利用复合形网提供的信息,从众多可行方向中提出了搜索网的伪梯度方向,从而使图网的柔韧性能增强,搜索方向更加明确,对一类约束非线性优化问题具有更高的计算效率.
Based on classical complex direct search method, by making good use of the information from complex network, the pseudo -gradient direction of a search network for constrained nonlinear optimization was put forward a-mong many feasible directions. Flexibility of complex network were increased and the direction of search network was more determined. Calculating results showed that there were faster search speed and calculating efficiency than complex search.
出处
《系统工程》
CSCD
1998年第2期17-23,共7页
Systems Engineering
关键词
伪梯度方向
直接搜索法
复合形法
非线性优化
Pseudo - Gradient Direction, Nonlinear Optimization, Direct Search, Complex Method, Constrained Optimization