LANS Publications
"Experiments with MINLP Branching Techniques"
S. Leyffer
Preprint ANL/MCS-P1734-0310
Preprint Version: [pdf]
Mixed-integer nonlinear optimization problems arise in scientific and operational applications ranging from the reordering of nuclear fuel rods to the design of wireless networks. We present some novel mixed-integer nonlinear optimization applications and review existing solution techniques. We also describe some experiments with nonlinear branch-and-bound branching techniques that lead us to promote a tighter integration of nonlinear solvers into a general branch-and-cut framework.