变量带上下界内点算法的理论与实现

Theory and Practice of Interior Point Algorithm with Upper-Lower Limit Variables

首先给出了变量带上下界的内点算法及其理论,并在计算机上实现了该算法。本文还提出了一个寻找初始内点的改进方法,数值计算表明改进后的方法能减少大多数情况下的迭代次数。

This paper introduces the interior point algorithm with upper-lower limit variables and its theory and practice. A modefied method in search of the initial interior point of the algorithm is presernted. The numerical calculation shows that the modified method may reduce the number of iterations under most circumstances.