不等式约束的凸二次规划问题的新算法

目的 研究求解不等式约束凸二次规划的新算法。方法 根据广义乘子法的思想，将具有不等式约束的凸二次规划问题转化为只有部分分量带非负约束的凸二次规划，通过解此简单凸二次规划问题建立凸二次规划的新算法。结果 新算法不用求逆矩阵，这样可充分保持矩阵的稀疏性，可用来解大规模稀疏问题。结论数值结果表明，在４８６／３３微机上就能解较大规模的凸二次规划。

New Algorithm of a Convex Quadratic program with Unequal Constraints

Aim To get new algorithm of a convex quadratic program with unequal constraints. Methods By the method of augmented lagrange multiplier, a convex quadratic program with unequal constraints was transformed into convex quadratic program with nonnegative constraints of part variables. New algorithm of convex quadratic program subject to unequal constraints was gotten by solving the problem. Results The new algorithm need not compute inverse matrix and then can solve large scale sparse convex quadratic programs. Conclusion The numerical results show the effectiveness of the new algorithm. It can solve larger scale sparse convex quadratic programs on 486/33 computer.