The DCAP Test Problems

Contributed by: Renan Garcia


This test problem suite consists of data and computational results for 12 two-stage stochastic integer programs arising in dynamic capacity acquisition and allocation applications. The problem formulation is described in [1]. The problems have mixed-integer first-stage variables, pure binary second-stage variables, and discrete distributions.

DATA:

The following table presents the sizes of the deterministic equivalents for each of the 12 problems instances and links to the data files. Columns NS, Rows, Cols, Bins, and Nonz refers to the number of scenarios, rows, columns, binary variables, and non-zero elements respectively.  The data for the problems is organized into 4  zipped files. Each file includes SMPS data for three problem instances corresponding to three scenario numbers (200,300,500). The core files are provided in both MPS as well as the CPLEX LP format (.lp extension).

____________________________________________________________________________

Name            NS          Size of the Det. Eq.          Data
                        Rows    Cols    Bins    Nonz               
dcap233_200     200     3012    5418    5406    10230   
dcap233_300     300     4512    8118    8106    15330   dcap233.zip
dcap233_500     500     7512    13518   13506   25530              
dcap243_200     200     3612    7218    7206    13230   
dcap243_300     300     5412    10818   10806   19830   dcap243.zip
dcap243_500     500     9012    18018   18006   33030                
dcap332_200     200     2412    4818    4806    9627    
dcap332_300     300     3612    7218    7206    14427   dcap332.zip
dcap332_500     500     6012    12018   12006   24027                

dcap342_200     200     2812    6418    6406    12427   
dcap342_300     300     4212    9618    9606    18627   dcap342.zip
dcap342_500     500     7012    16018   16006   31027______________
SOLUTION:

The following table presents the best known upper and lower bounds obtained with a specialized decomposition based branch & bound (DBB) algorithm (described in [1] and [2]), and that obtained by solving the deterministic equivalent MIP using the CPLEX 7.5 MIP solver (with default options). A CPU time limit of 1800 secs and a termination tolerance of 0.01% relative optimality gap was used. The computations were performed on a Sun Ultra 60 workstation with 512 MB RAM and a 480MHz processor.

____________________________________________________________________________________________________________________________
             |                   DBB Algorithm                      |                   CPLEX 7.5                                                  
             |   Iter    Nodes   LB        UB        % Gap   CPUs   |  Iter    Nodes   LB         UB         % Gap   CPUs   
dcap233_200  |   343     687     1834.868  1834.893  0.0014  18.15  |  2263792 92397   1823.8912  1854.8395  1.6968  1802.68
dcap233_300  |   128     257     1644.342  1644.504  0.0098  10.25  |  1476305 19222   1623.1511  1657.1376  2.0939  1801.00
dcap233_500  |   124     249     1737.719  1737.728  0.0005  15.89  |  621088  36565   1712.0552  1748.6441  2.1371  1802.48
dcap243_200  |   331     663     2322.721  2322.721  0.0000  51.68  |  468340  103803  2299.1849  2352.5774  2.3222  1802.95
dcap243_300  |   100     201     2559.297  2559.545  0.0097  23.48  |  190918  73342   2541.5712  2640.6479  3.8982  1802.47
dcap243_500  |   231     463     2167.512  2167.512  0.0000  83.84  |  420752  21677   2128.5289  2197.3689  3.2342  1801.93
dcap332_200  |   1754    3509    1060.676  1060.781  0.0100  133.34 |  350650  151932  1059.4573  1061.1722  0.1619  1802.63
dcap332_300  |   1270    2541    1253.040  1253.082  0.0033  141.32 |  320464  94357   1251.0195  1257.7704  0.5396  1803.20
dcap332_500  |   1155    2311    1588.717  1588.876  0.0100  199.67 |  765243  35132   1587.6687  1590.7464  0.1939  1801.93
dcap342_200  |   462     925     1619.531  1619.654  0.0076  131.94 |  390094  113059  1619.2613  1620.9507  0.1043  1802.62
dcap342_300  |   980     1961    2067.777  2067.818  0.0020  439.15 |  429196  59446   2066.9391  2071.3888  0.2153  1802.69
dcap342_500  |   510     1021    1904.722  1904.803  0.0043  349.98 |  679448  14831   1903.7113  1913.4352  0.5108  1801.23


REFERENCES:

[1] S. Ahmed and R. Garcia. "Dynamic Capacity Acquisition and Assignment under Uncertainty," Annals of Operations Research, vol.124, pp. 267-283, 2003.

[2] S. Ahmed, M. Tawarmalani, and N. V. Sahinidis. "A Finite Branch and Bound Algorithm for Two-stage Stochastic Integer Programs,'' Mathematical Programming , vol.100, pp.355-377, 2004.



[SIPLIB HOME]