Comparison of Quadratic Convex Reformulations to Solve the Quadratic Assignment Problem
We consider the (QAP) that consists in minimizing a quadratic function subject to assignment constraints where the variables are binary. In this paper, we build two families of equivalent quadratic convex formulations of (QAP). The continuous relaxation of each equivalent formulation is then a convex problem and can be used within a B&B. In this work, we focus on finding the “best” equivalent formulation within each family, and we prove that it can be computed using semidefinite programming. Finally, we get two convex formulations of (QAP) that differ from their sizes and from the tightness of their continuous relaxation bound. We present computational experiments that prove the practical usefulness of using quadratic convex formulation to solve instances of (QAP) of medium sizes.
KeywordsQuadratic assignment problem Convex quadratic programming Semidefinite programming Experiments
- 4.Billionnet, A., Elloumi, S., Lambert, A., Wiegele, A.: Using a conic bundle method to accelerate both phases of a quadratic convex reformulation. Inf. J. Comput. (2016, to appear)Google Scholar
- 6.Burkard, R.E.: Quadratic Assignment Problems, pp. 2741–2814. Springer, New York (2013)Google Scholar
- 8.Helmberg, C.: Conic Bundle v0.3.10 (2011)Google Scholar
- 9.IBM-ILOG: IBM ILOG CPLEX 12.6 Reference Manual (2014). http://www-01.ibm.com/support/knowledgecenter/SSSA5P_12.6.0/ilog.odms.studio.help/Optimization_Studio/topics/COS_home.html