Solution Methods for Stochastic Dynamic Linear Programs.
Abstract
Linear programs have been formulated for many practical situations that require decisions made periodically through time. These dynamic linear programs often involve uncertainties. Deterministic solutions of these problems may lead to costly incorrect decisions, and, when a stochastic solution is attempted the problem may become too large. In this report, we present methods for reducing the computational cost of these stochastic programs, and we show conditions under which the stochastic program need not be solved. Our methods are based on the large-scale programming techniques of decomposition, partitioning, and basic factorization. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1980
- Accession Number
- ADA096119
Entities
People
- John R. Birge
Organizations
- Stanford University