Application of Optimizied Fuzzy System by Genetic Algorithm to Predict Air Demand Downstream of Bottom Outlet Gates

Bottom outlet conduits are one of the main components of high dams to control the reservoir capacity,
sediment discharge and downstream requirements. With high velocity flows and discharges within the
bottom outlet conduits, the potential for cavitation damage, especially downstream of the outlet gates
is also increased. Amongst various methods, aeration has been found as a cheap and easy way to
reduce the cavitation tendency. Based on complicated phenomena of two phase flow downstream of
outlet gates, the estimation of air demand without model studies mainly results in substantial errors.
On the other hand, in recent years the application of artificial intelligence in the forms of artificial
neural networks and fuzzy logic have been found as a powerful technique for simulation of nonlinear
systems. A combination of fuzzy systems and genetic algorithms and optimization based on
experimental data is an effective way for nonlinear problems. In this paper, a model based on geneticfuzzy
algorithm has been developed and used to predict air demand downstream of bottom outlet gates
of high dams. The model make uses of 243 data consisting the data collected from various hydraulic
models of outlet dams in Iran and the data of Folsom dam in USA. The model shows satisfactory
results and proves to be a reliable technique to predict air demand downstream of bottom outlet gates,
compared with the various relationships which are developed by regression analysis of experimental
and field data.