An efficient algorithm for globally minimizing sum of quadratic ratios problem with nonconvex quadratic constraints (Q2383678): Difference between revisions

A new branch and bound algorithm is proposed for solving the following minimization problem: \[ \text{Minimize }\sum^p_{j=1}h_j (t)=\sum^p_{j=1}\gamma_j (n_j(t)/d_j(t)) \] \[ \text{subject to } g_m(t)\leq 0,\;m=1,\dots,M,\;t^l\leq t\leq t^u, \] where \(t\in\mathbb{R}^N\), and \(n_j(t)\), \(d_i(t)\), \(g_m(t)\) are non-convex quadratic functions. In the proposed algorithm, the interval Newton method is used to facilitate convergence in the neighborhood of the global solution. The algorithm is convergent to the global optimum through the successive refinement of the solutions of a series of linear programming problems. Numerical experiments demonstrate the efficiency of the proposed algorithm.