无严格互补松驰条件的序列线性方程组新算法

该文通过构造特殊形式的有效集来逼近KKT点处的有效集，给出了一个任意初始点下的序列线性方程组新算法，并证明了该算法在没有严格互补松驰条件的情况下具有全局收敛性和一步超线性收敛性。 

A New Sequential Systems of Linear Equations Algorithm Without Strict Complementary Slackness

In this paper, a new algorithm of sequential systems of linear equations (SSLE) with arbitrary initial point is proposed by constructing a special active constraints set. Only one or two systems of linear equations with the same coefficient matrix are solved per iteration. The algorithm is globally convergent. It worths to point out that the algorithm still has one -step superlinear convergence under some weaker assumptions, for instance, without the strict complementary slackness condition.