摘要
基于牛顿方向,给出了求解凸二次规划问题的改进原对偶可行内点算法。若获得算法的初始可行内点,则该算法经过多次迭代之后收敛到原问题的一个最优解。数值试验表明了该算法的有效性。
In this paper we analyzed the most common algorithms to quadratic programming and indicated difficulty in studying this problem.Based on above,we presented an improved primal-dual feasible interior point algorithm for convex quadratic programming by means of the Newton direction.It is showed that if a strictly feasible starting point is available,then the algorithms have the polynomial complexity.Numerical results are demonstrated very good computational performance on convex quadratic programming.
出处
《长江大学学报（自科版）（上旬）》
CAS
2009年第2期126-128,共3页
JOURNAL OF YANGTZE UNIVERSITY (NATURAL SCIENCE EDITION) SCI ＆ ENG
关键词
凸二次规划
原对偶可行内点算法
多项式复杂性
convex quadratic programming
primal-dual interior point algorithm
polynomial complexity
