Author(s): C. A. S. Farias; E. C. M. Machado; L. N. Brasiliano
Linked Author(s):
Keywords: No keywords
Abstract: This paper reports on the application of Implicit Stochastic Optimization (ISO) and SelfOrganizing Maps (SOM) to determine monthly operating policies for a reservoir system located in semiarid Brazil. Three steps define the ISO-SOM model: (1) generation of monthly inflow scenarios; (2) determinist optimization of the system operation under the generated inflow scenarios; and (3) development of monthly rules by means of SOM, which consists of conditioning allocation coefficients on initial reservoir storage and current inflow. The study demonstrates the superiority of ISO-SOM model over the standard rules of reservoir operation, especially during drought spells.
Year: 2016