Abstract:Deterministic nonconvex optimization solvers generate convex relaxations of nonconvex functions by making use of underlying factorable representations. One approach introduces auxiliary variables assigned to each factor that lifts the problem into a higher-dimensional decision space. In contrast, a generalized McCormick relaxation approach offers the significant advantage of constructing relaxations in the lower dimensionality space of the original problem without introducing auxiliary variables, often referred to as a "reduced-space" approach. Recent contributions illustrated how additional nontrivial inequality constraints may be used in factorable programming to tighten relaxations of the ubiquitous bilinear term. In this work, we exploit an analogous representation of McCormick relaxations and factorable programming to formulate tighter relaxations in the original decision space. We develop the underlying theory to generate necessarily tighter reduced-space McCormick relaxations when a priori convex/concave relaxations are known for intermediate bilinear terms. We then show how these rules can be generalized within a McCormick relaxation scheme via three different approaches: the use of a McCormick relaxations coupled to affine arithmetic, the propagation of affine relaxations implied by subgradients, and an enumerative approach that directly uses relaxations of each factor. The developed approaches are benchmarked on a library of optimization problems using the EAGO.jl optimizer. Two case studies are also considered to demonstrate the developments: an application in advanced manufacturing to optimize supply chain quality metrics and a global dynamic optimization application for rigorous model validation of a kinetic mechanism. The presented subgradient method leads to an improvement in CPU time required to solve the considered problems to -global optimality.

Complexity of linear relaxations in integer programming

Convex quadratic sets and the complexity of mixed integer convex quadratic programming

Convex Relaxation and Lagrangian Decomposition for Indefinite Integer Quadratic Programming

MIP Relaxations in Factorable Programming

The Subspace Flatness Conjecture and Faster Integer Programming

Relaxations and cutting planes for linear programs with complementarity constraints

Parametric Integer Programming in Fixed Dimension

New finite relaxation hierarchies for concavo-convex, disjoint bilinear programs, and facial disjunctions

A Slightly Lifted Convex Relaxation for Nonconvex Quadratic Programming with Ball Constraints

On exact and inexact RLT and SDP-RLT relaxations of quadratic programs with box constraints

Improved Convex and Concave Relaxations of Composite Bilinear Forms

Statistical Limits of Convex Relaxations

On the strength of Burer's lifted convex relaxation to quadratic programming with ball constraints

Strengthening Convex Relaxations of 0/1-Sets Using Boolean Formulas

On the strength of recursive McCormick relaxations for binary polynomial optimization

Instance-specific linear relaxations of semidefinite optimization problems

Linear Programming Relaxations of Quadratically Constrained Quadratic Programs

On Tractable Convex Relaxations of Standard Quadratic Optimization Problems under Sparsity Constraints

On Connections Between Association Schemes and Analyses of Polyhedral and Positive Semidefinite Lift-and-Project Relaxations

Lower bounds and integrality gaps in simplicial decomposition