On convex relaxations of quadrilinear terms

被引:31
作者
Cafieri, Sonia [1 ]
Lee, Jon [2 ]
Liberti, Leo [1 ]
机构
[1] Ecole Polytech, LIX, F-91128 Palaiseau, France
[2] IBM Corp, Thomas J Watson Res Ctr, Dept Math Sci, Yorktown Hts, NY 10598 USA
关键词
Quadrilinear; Trilinear; Bilinear; Convex relaxation; Reformulation; Global optimization; Spatial branch and bound; MINLP; GLOBAL OPTIMIZATION; C-2-CONTINUOUS PROBLEMS; TRILINEAR MONOMIALS; UNDERESTIMATORS; ENVELOPES; DOMAINS; FACETS; BOUNDS;
D O I
10.1007/s10898-009-9484-1
中图分类号
C93 [管理学]; O22 [运筹学];
学科分类号
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要
The best known method to find exact or at least epsilon-approximate solutions to polynomial-programming problems is the spatial Branch-and-Bound algorithm, which rests on computing lower bounds to the value of the objective function to be minimized on each region that it explores. These lower bounds are often computed by solving convex relaxations of the original program. Although convex envelopes are explicitly known (via linear inequalities) for bilinear and trilinear terms on arbitrary boxes, such a description is unknown, in general, for multilinear terms of higher order. In this paper, we study convex relaxations of quadrilinear terms. We exploit associativity to rewrite such terms as products of bilinear and trilinear terms. Using a general technique, we formally establish the intuitive fact that any relaxation for k-linear terms that employs a successive use of relaxing bilinear terms (via the bilinear convex envelope) can be improved by employing instead a relaxation of a trilinear term (via the trilinear convex envelope). We present a computational analysis which helps establish which relaxations are strictly tighter, and we apply our findings to two well-studied applications: the Molecular Distance Geometry Problem and the Hartree-Fock Problem.
引用
收藏
页码:661 / 685
页数:25
相关论文
共 41 条
  • [1] A global optimization method, αBB, for general twice-differentiable constrained NLPs -: I.: Theoretical advances
    Adjiman, CS
    Dallwig, S
    Floudas, CA
    Neumaier, A
    [J]. COMPUTERS & CHEMICAL ENGINEERING, 1998, 22 (09) : 1137 - 1158
  • [2] ADJIMAN CS, 1998, THESIS PRINCETON U
  • [3] JOINTLY CONSTRAINED BICONVEX PROGRAMMING
    ALKHAYYAL, FA
    FALK, JE
    [J]. MATHEMATICS OF OPERATIONS RESEARCH, 1983, 8 (02) : 273 - 286
  • [4] [Anonymous], **NON-TRADITIONAL**
  • [5] Avis David., lrs
  • [6] Branching and bounds tightening techniques for non-convex MINLP
    Belotti, Pietro
    Lee, Jon
    Liberti, Leo
    Margot, Francois
    Waechter, Andreas
    [J]. OPTIMIZATION METHODS & SOFTWARE, 2009, 24 (4-5) : 597 - 634
  • [7] Crippen G.M., 1988, Distance Geometry and Molecular Conformation, Volume15 of Chemometrics Series
  • [8] FLOUDAS C, 2007, COMMUNICATION
  • [9] Fourer R., 2002, AMPL BOOK
  • [10] Fukuda K., cdd