Multi-Reservoir Systems Optimization by the Aggregation/Decomposition Methodology

by M. Bessa, Univ of Waterloo, Waterloo, Canada,
K. Ponnambalam, Univ of Waterloo, Waterloo, Canada,
A. Bogobowicz, Univ of Waterloo, Waterloo, Canada,

Document Type: Proceeding Paper

Part of: Hydraulic Engineering


Studies on long term optimization of multi-reservoir systems are enhanced by the availability of closed-loop policies as available in Stochastic Dynamic Programming (SDP). SDP offers a great deal of desired characteristics but, when the number of reservoir exceeds three, it is nearly impractical to solve the problem. One way to overcome this hindrance is to make use of approximate models using, for example, an aggregation/decomposition methodology. The reduced model makes it feasible to solve multi-reservoir stochastic optimization problems using SDP. However, because of the aggregation/decomposition methodology, the policies determined are only sub optimal. In this paper, this new methodology of optimization for multi-reservoir problems is presented with an application to a four-reservoir operational problem.

Subject Headings: Control systems | Hydrologic models | Water supply systems | Reservoirs | Computer programming | Flow control | Water flow | Optimization models

Services: Buy this book/Buy this article


Return to search