An accelerating algorithm for globally solving nonconvex quadratic programming

To globally solve a nonconvex quadratic programming problem, this paper presents an accelerating linearizing algorithm based on the framework of the branch-and-bound method. By utilizing a new linear relaxation approach, the initial quadratic programming problem is reduced to a sequence of linear relaxation programming problems, which is used to obtain a lower bound of the optimal value of this problem. Then, by using the deleting operation of the investigated regions, we can improve the convergent speed of the proposed algorithm. The proposed algorithm is proved to be convergent, and some experiments are reported to show higher feasibility and efficiency of the proposed algorithm.