ABS算法在线性约束优化中的应用

APPLICATIONS OF ABS ALGORITHMS TO LINEAR CONSTRAINED OPTIMIZATION

本文介绍ABS算法在线性约束优化中的应用,利用ABS算法对GI法进行了重新描述,给出了投影矩阵及Moore-Penrose伪逆连续修正的ABS方法,并探讨了利用ABS算法来构造下降方向的方法。

In this paper, we present the applications of ABS algorithms to linear constrained optimization, reformulate the GI method via ABS algorithms without use of Givens transformations, give the methods for updating projection metrices and Moore-Penrose pseudo invers successively, and discuss the ways to construct descent direction by ABS algorithms.