A parallel implementation of the nested decomposition algorithm for multistage stochastic linear programs

Multistage stochastic linear programs can represent a variety of practical decision problems. Solving a multistage stochastic program can be viewed as solving a large tree of linear programs. A common approach for solving these problems is the nested decomposition algorithm, which moves up down the tree by solving nodes and passing information among nodes. The natural independence of subtrees suggests that much of the computational effort of the nested decomposition algorithm can run in parallel across small numbers of fast processors. This paper explores the advantages of such parallel implementations over serial implementations and compares alternative sequencing protocols for parallel processors. Computational experience on a large test set of practical problems with up to 1.5 million constraints and almost 5 million variables suggests that parallel implementations may indeed work well, but they require careful attention to processor load balancing.

[1]  Alan J. King,et al.  A Standard Input Format for Multiperiod Stochastic Linear Programs , 1987 .

[2]  Andrzej Ruszczynski,et al.  Parallel decomposition of multistage stochastic programming problems , 1993, Math. Program..

[3]  John M. Mulvey,et al.  A New Scenario Decomposition Method for Large-Scale Stochastic Optimization , 1995, Oper. Res..

[4]  John R. Birge,et al.  Efficient solution of two-stage stochastic linear programs using interior point methods , 1992, Comput. Optim. Appl..

[5]  Andrzej Ruszczynski,et al.  A regularized decomposition method for minimizing a sum of polyhedral functions , 1986, Math. Program..

[6]  David P. Rutenberg,et al.  Computation in Discrete Stochastic Programs with Recourse , 1973, Oper. Res..

[7]  R. J-B. Wets,et al.  Large Scale Linear Programming Techniques , 1988 .

[8]  J. Mulvey,et al.  Stochastic network optimization models for investment planning , 1989 .

[9]  Horand I. Gassmann,et al.  Optimal harvest of a forest in the presence of uncertainty , 1989 .

[10]  James K. Ho,et al.  A set of staircase linear programming test problems , 1981, Math. Program..

[11]  Horand I. Gassmann,et al.  Mslip: A computer code for the multistage stochastic linear programming problem , 1990, Math. Program..

[12]  John R. Birge,et al.  Decomposition and Partitioning Methods for Multistage Stochastic Linear Programs , 1985, Oper. Res..

[13]  R. Tyrrell Rockafellar,et al.  Scenarios and Policy Aggregation in Optimization Under Uncertainty , 1991, Math. Oper. Res..

[14]  K. A. Ariyawansa,et al.  Performance of a benchmark parallel implementation of the Van Slyke and Wets algorithm for two-stage stochastic programs on the Sequent/Balance , 1991, Concurr. Pract. Exp..

[15]  Stavros A. Zenios,et al.  A Massively Parallel Algorithm for Nonlinear Stochastic Network Problems , 1993, Oper. Res..

[16]  Robert Entriken.  A Parallel Decomposition Algorithm for Staircase Linear Programs , 1988 .

[17]  R. Wets,et al.  L-SHAPED LINEAR PROGRAMS WITH APPLICATIONS TO OPTIMAL CONTROL AND STOCHASTIC PROGRAMMING. , 1969 .

[18]  J. Birge,et al.  A multicut algorithm for two-stage stochastic linear programs , 1988 .

[19]  Robert J. Wittrock.  Dual nested decomposition of staircase linear programs , 1985 .

[20]  Roger J.-B. Wets,et al.  Lifting projections of convex polyhedra. , 1969 .