Premium
Evolutionary computation to control unnatural water level fluctuations in multi‐reservoir river systems
Author(s) -
Dessalegne Tibebe,
Nicklow John,
Minder Eric
Publication year - 2004
Publication title -
river research and applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.679
H-Index - 94
eISSN - 1535-1467
pISSN - 1535-1459
DOI - 10.1002/rra.793
Subject(s) - genetic algorithm , computation , computer science , water level , hydrology (agriculture) , mathematical optimization , flow (mathematics) , environmental science , algorithm , geology , mathematics , geotechnical engineering , geometry , cartography , geography
Abstract A methodology and computational model for evaluation and optimization of dam operations in multi‐reservoir river systems is presented. The optimal dam operation scheme was defined as that which minimizes unnatural water level fluctuations during low flow periods, subject to the maintenance of desired stages within the river network, desired storage levels in the reservoirs, and bounds on releases from dams. The dam operation problem was solved using a discrete‐time optimal control methodology in which a genetic algorithm was interfaced with an existing, unsteady, one‐dimensional hydraulic simulation model known as FLDWAV. The hydraulic simulation model was used to solve governing hydraulic constraints, while the genetic algorithm was employed to solve the overall control problem. Constraints on water surface and storage levels were handled using an Additive Penalty Method. Two types of objective functions, absolute water surface deviation and rate of water surface fluctuation, were considered and their results were compared. Capabilities of the overall model were demonstrated through applications to a hypothetical three‐reservoir river system and a portion of the Illinois River Waterway, for which the hydraulic simulation model was calibrated and verified. Copyright © 2004 John Wiley & Sons, Ltd.