TY - JOUR
T1 - A branch-and-cut algorithm without binary variables for nonconvex piecewise linear optimization
AU - Keha, Ahmet B.
AU - De Farias, Ismael R.
AU - Nemhauser, George L.
PY - 2006/9
Y1 - 2006/9
N2 - We give a branch-and-cut algorithm for solving linear programs (LPs) with continuous separable piecewise-linear cost functions (PLFs). Models for PLFs use continuous variables in special-ordered sets of type 2 (SOS2). Traditionally, SOS2 constraints are enforced by introducing auxiliary binary variables and other linear constraints on them. Alternatively, we can enforce SOS2 constraints by branching on them, thus dispensing with auxiliary binary variables. We explore this approach further by studying the inequality description of the convex hull of the feasible set of LPs with PLFs in the space of the continuous variables, and using the new cuts in a branch-and-cut scheme without auxiliary binary variables. We give two families of valid inequalities. The first family is obtained by lifting the convexity constraints. The second family consists of lifted cover inequalities. Finally, we report computational results that demonstrate the effectiveness of our cuts, and that branch-and-cut without auxiliary binary variables is significantly more practical than the traditional mixed-integer programming approach.
AB - We give a branch-and-cut algorithm for solving linear programs (LPs) with continuous separable piecewise-linear cost functions (PLFs). Models for PLFs use continuous variables in special-ordered sets of type 2 (SOS2). Traditionally, SOS2 constraints are enforced by introducing auxiliary binary variables and other linear constraints on them. Alternatively, we can enforce SOS2 constraints by branching on them, thus dispensing with auxiliary binary variables. We explore this approach further by studying the inequality description of the convex hull of the feasible set of LPs with PLFs in the space of the continuous variables, and using the new cuts in a branch-and-cut scheme without auxiliary binary variables. We give two families of valid inequalities. The first family is obtained by lifting the convexity constraints. The second family consists of lifted cover inequalities. Finally, we report computational results that demonstrate the effectiveness of our cuts, and that branch-and-cut without auxiliary binary variables is significantly more practical than the traditional mixed-integer programming approach.
KW - Branch-and-cut
KW - Piecewise linear function
KW - Special-ordered set
UR - http://www.scopus.com/inward/record.url?scp=33749636873&partnerID=8YFLogxK
U2 - 10.1287/opre.1060.0277
DO - 10.1287/opre.1060.0277
M3 - Article
AN - SCOPUS:33749636873
SN - 0030-364X
VL - 54
SP - 847
EP - 858
JO - Operations Research
JF - Operations Research
IS - 5
ER -