Tighter Convex Relaxations for Global Optimization Using alphaBB Based Approach

This paper is devoted to investigation of certain issues that appear in solving of deterministic global optimization problems (GOPs). Basically, we focus ourselves on introducing a procedure which may serve to establish tighter convex relaxations for a certain class of non-convex optimization problems. Tightness of these convex relaxations plays important role in speeding of the convergence of branch-and-bound algorithm which is used as a basic framework of solving GOPs in this study. Two case studies are solved where it is shown how significant improvement can be achieved by considering proposed framework.