A QUASI-SIMPLEX METHOD FOR STRICTLY CONVEX QUADRATIC PROGRAMMING

本文提出一个不用 Kuhn- Tucker条件而直接搜索严格凸二次规划最优目标点的鲁棒方法 .在搜索过程中 ,目标点沿约束多面体边界上的一条折线移动 .这种移动目标点的思想可以被认为是线性规划单纯形法的自然推广 ,在单纯形法中 ,目标点从一个顶点移到另一个顶点。

In this paper, a robust method of directly searching the optimum objective point for strictly convex quadratic programming is presented, while no Kuhn-Tucker condition is applied to the solution. During the searching procedure the objective point is moved along a broken line on the boundary of the constraint polyhedron. This idea moving the objective point may be considered a natural generalization of the simplex method for linear programming where the objective point is moved from a vertex to another.

辽宁省自然科学基金部分资助