Conflict analysis in mixed integer programming

被引:63
作者
Achterberg, Tobias [1 ]
机构
[1] Konrad Zuse Zentrum Informat Tech, Berlin, Germany
关键词
mixed integer programming; branch-and-cut; conflict analysis; SAT; infeasible mixed integer programming;
D O I
10.1016/j.disopt.2006.10.006
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
Conflict analysis for infeasible subproblems is one of the key ingredients in modern SAT solvers. III contrast, it is common practice for today's mixed integer programming solvers to discard infeasible subproblems and the information they reveal. III this paper, we try to remedy this situation by generalizing SAT infeasibility analysis to mixed integer programming. We present heuristics for branch-and-cut solvers to generate valid inequalities from the Current infeasible subproblem and the associated branching information. SAT techniques can then be used to strengthen the resulting constraints. Extensive Computational experiments show the potential Of Our method. Conflict analysis greatly improves the performance oil particular models, while it does not interfere with the solving process on the other instances. In total, the number of required branching nodes oil general MIP instances was reduced by 18% in the geometric mean, and the solving time was reduced by 11%. On infeasible MIP arising in the context of chip verification and on a model for a particular combinatorial game, the number of nodes was reduced by 80%, thereby reducing the solving time by 50%. (c) 2006 Elsevier B.V. All rights reserved.
引用
收藏
页码:4 / 20
页数:17
相关论文
共 28 条
  • [1] ACHTERBERG T, 2003, MIXED INTEGER PROGRA
  • [2] ACHTERBERG T, 0419 ZUS I
  • [3] ACHTERBERG T, 2006, OPER RES LETT, V34, P1, DOI DOI 10.1016/J.ORL.2005.07.009
  • [4] On the maximum feasible subsystem problem, IISs and IIS-hypergraphs
    Amaldi, E
    Pfetsch, ME
    Trotter, LE
    [J]. MATHEMATICAL PROGRAMMING, 2003, 95 (03) : 533 - 554
  • [5] Conflict graphs in solving integer programming problems
    Atamtürk, A
    Nemhauser, GL
    Savelsbergh, MWP
    [J]. EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2000, 121 (01) : 40 - 55
  • [6] BIERE A, 2002, ACM IEEE INT C COMP
  • [7] FINITENESS PROOF FOR MODIFIED DANTZIG CUTS IN INTEGER PROGRAMMING
    BOWMAN, VJ
    NEMHAUSE.GL
    [J]. NAVAL RESEARCH LOGISTICS QUARTERLY, 1970, 17 (03): : 309 - &
  • [8] RTL-datapath verification using integer linear programming
    Brinkmann, R
    Drechsler, R
    [J]. ASP-DAC/VLSI DESIGN 2002: 7TH ASIA AND SOUTH PACIFIC DESIGN AUTOMATION CONFERENCE AND 15TH INTERNATIONAL CONFERENCE ON VLSI DESIGN, PROCEEDINGS, 2002, : 741 - 746
  • [9] Cook S.A., 1971, P 3 ANN ACM S THEOR, P151, DOI DOI 10.1145/800157.805047
  • [10] Exploring relaxation induced neighborhoods to improve MIP solutions
    Danna, E
    Rothberg, E
    Le Pape, C
    [J]. MATHEMATICAL PROGRAMMING, 2005, 102 (01) : 71 - 90