A.M. Iliev, V.V. Fustik, and I. Kukovski (Macedonia)
Modeling, Hydro unit, ANFIS, Fuzzy logic, Efficiency.
In this paper a method for calculation of the hydro unit efficiency using adaptive network-based fuzzy inference system (ANFIS) is presented. The essential characteristic of the ANFIS is that the fuzzy system learns from the data that are used for modeling. Thus, the parameters associ ated with the membership functions vary through the learning process in order to perform an accurate model. Input variables of this system are the net head and the water flow through hydro turbine, while the output vari able is the efficiency of the hydro unit. The modeling of those variables is realized by the first order Sugeno-type system. The parameters of membership functions are tuned using a combination of backpropagation and the least square method. Validation of the developed model is done with a real set of training and checking data obtained from the measurement in Hydro Power Plant Tikves. Computer simulation is performed with MatLab Fuzzy Logic Toolbox.
Important Links:
Go Back
