Application of the Bernstein form for reliable global optimization of mixed-integer problems in process synthesis and design

In this paper, we present a Bernstein polynomial approach to optimally and reliably solve polynomial class of mixed-integer programming (MIP) problems. We use the Bernstein form in the branch-and-bound framework to compute a global optimum of MIPs. We present two application examples of MIPs: optimal pump configuration problem and trim loss minimization, to demonstrate the Bernstein polynomial approach.