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 genetic-fuzzy 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.