Algorithms For Bound Constrained Convex Quadratic Programming

This paper studies the methods for bounded constrained convex quadraticprogramming. In well-known SQP methods for nonlinear programming, the search direction isthe solution of a bounded constrained convex quadratic programming. We considerstrict convex quadratic programming and positive semidefinite convex quadraticprogramming. For the former, this paper proposed the modified method whichcombines regular splitting and projected search. The convergence of modified methodis proved. For the latter, this paper combines Cholesky decomposition and branch andbound, and develops a method for positive semidefinite convex quadraticprogramming. The convergence of this method is proved. In the last part of this paper, numerical results for the two methods are givenand these results show that two methods are effective.