Projected Perspective Reformulations for MIQP problems

The Perspective Relaxation (PR) is a general approach for constructing tight approximations to Mixed-Integer NonLinear Problems with semicontinuous variables. The PR of a MINLP can be formulated either as a Mixed-Integer Second-Order Cone Program, provided that the original objective function is SOCP-representable, or as a Semi-Infinite MINLP. We show that under some further assumptions—rather restrictive, but satisfied in several practical applications—the PR of Mixed-Integer Quadratic Programs can also be reformulated as a piecewise linear-quadratic problem of roughly the same size of the standard continuous relaxation. Furthermore, if the original problem has some exploitable structure, this is typically preserved in the reformulation, allowing to construct specialized approaches for solving the PR. We report on implementing these ideas on two MIQPs with appropriate structure: a sensor placement problem and a Quadratic-cost (single-commodity) network design problem.

[1]  Oktay Günlük,et al.  IBM Research Report MINLP Strengthening for Separable Convex Quadratic Transportation-Cost UFL , 2007 .

[2]  Jeff T. Linderoth,et al.  RELAXATION OF MINLPS WITH INDICATOR VARIABLES , 2007 .

[3]  Claudio Gentile,et al.  SDP diagonalizations and perspective cuts for a class of nonseparable MIQP , 2007, Oper. Res. Lett..

[4]  Claudio Gentile,et al.  Perspective cuts for 0-1 mixed integer programs , 2002 .

[5]  Alessandro Agnetis,et al.  Covering a line segment with variable radius discs , 2007, Comput. Oper. Res..

[6]  J. Castro,et al.  An implementation of linear and nonlinear multicommodity network flows , 1996 .

[7]  Darwin Klingman,et al.  NETGEN: A Program for Generating Large Scale Capacitated Assignment, Transportation, and Minimum Cost Flow Network Problems , 1974 .

[8]  Robert R. Meyer,et al.  A Partitioned ∈-Relaxation Algorithm for Separable Convex Network Flow Problems , 1999, Comput. Optim. Appl..

[9]  Alessandro Agnetis,et al.  Demand allocation with latency cost functions , 2008, Math. Program..

[10]  Nikolaos V. Sahinidis,et al.  Convex extensions and envelopes of lower semi-continuous functions , 2002, Math. Program..

[11]  Sanjay Mehrotra,et al.  A branch-and-cut method for 0-1 mixed convex programming , 1999, Math. Program..

[12]  C. Gentile,et al.  Tighter Approximated MILP Formulations for Unit Commitment Problems , 2009, IEEE Transactions on Power Systems.

[13]  Sinan Gürel,et al.  A strong conic quadratic reformulation for machine-job assignment with controllable processing times , 2009, Oper. Res. Lett..

[14]  R. Meyer,et al.  A Partitioned ∈-Relaxation Algorithm for Separable Convex Network Flow Problems , 1999, Comput. Optim. Appl..

[15]  Sebastián Ceria,et al.  Convex programming for disjunctive convex optimization , 1999, Math. Program..

[16]  Claudio Gentile,et al.  A computational comparison of reformulations of the perspective relaxation: SOCP vs. cutting planes , 2009, Oper. Res. Lett..