Exploring relaxation induced neighborhoods to improve MIP solutions

被引:297
作者
Danna, E
Rothberg, E
Le Pape, C
机构
[1] ILOG SA, F-94253 Gentilly, France
[2] CNRS, FRE 2487, Lab Informat Avignon, F-84911 Avignon 9, France
[3] ILOG Inc, Mountain View, CA 94043 USA
关键词
mixed integer programming; heuristics; local search; hybrid algorithms;
D O I
10.1007/s10107-004-0518-7
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Given a feasible solution to a Mixed Integer Programming (MIP) model, a natural question is whether that solution can be improved using local search techniques. Local search has been applied very successfully in a variety of other combinatorial optimization domains. Unfortunately, local search relies extensively on the notion of a solution neighborhood, and this neighborhood is almost always tailored to the structure of the particular problem being solved. A MIP model typically conveys little information about the underlying problem structure. This paper considers two new approaches to exploring interesting, domain-independent neighborhoods in MIP. The more effective of the two, which we call Relaxation Induced Neighborhood Search (RINS), constructs a promising neighborhood using information contained in the continuous relaxation of the MIP model. Neighborhood exploration is then formulated as a MIP model itself and solved recursively. The second, which we call guided dives, is a simple modification of the MIP tree traversal order. Loosely speaking, it guides the search towards nodes that are close neighbors of the best known feasible solution. Extensive computational experiments on very difficult MIP models show that both approaches outperform default CPLEX MIP and a previously described approach for exploring MIP neighborhoods (local branching) with respect to several different metrics. The metrics we consider are quality of the best integer solution produced within a time limit, ability to improve a given integer solution (of both good and poor quality), and time required to diversify the search in order to find a new solution.
引用
收藏
页码:71 / 90
页数:20
相关论文
共 35 条
[1]  
Aboudi R., 1994, ORSA Journal on Computing, V6, P82, DOI 10.1287/ijoc.6.1.82
[2]   THE SHIFTING BOTTLENECK PROCEDURE FOR JOB SHOP SCHEDULING [J].
ADAMS, J ;
BALAS, E ;
ZAWACK, D .
MANAGEMENT SCIENCE, 1988, 34 (03) :391-401
[3]   RECENT ADVANCES IN CREW-PAIRING OPTIMIZATION AT AMERICAN-AIRLINES [J].
ANBIL, R ;
GELMAN, E ;
PATTY, B ;
TANGA, R .
INTERFACES, 1991, 21 (01) :62-74
[4]  
Applegate D., 1991, ORSA Journal on Computing, V3, P149, DOI 10.1287/ijoc.3.2.149
[5]   PIVOT AND COMPLEMENT - A HEURISTIC FOR 0-1 PROGRAMMING [J].
BALAS, E ;
MARTIN, CH .
MANAGEMENT SCIENCE, 1980, 26 (01) :86-96
[6]   Octane: A new heuristic for pure 0-1 programs [J].
Balas, E ;
Ceria, S ;
Dawande, M ;
Margot, F ;
Pataki, G .
OPERATIONS RESEARCH, 2001, 49 (02) :207-225
[7]  
BAPTISTE P, 1995, P 1 INT JOINT WORKSH
[8]  
Bixby RE, 1998, Optima, V58, P12
[9]  
BIXBY RE, 2000, MIP THEORY PRACTICE, P19
[10]  
CASEAU Y, 1997, LIENS9711 EC NORM SU