sajjad Naderian

Most hydrological series due to the complexity of the physical and temporal and spatial variations, are not static. In such circumstances, the use of artificial intelligence models will be very useful. One of the ways which in recent years have been taken into consideration in the field of hydrology is the use of hybrid model of wavelet-ANFIS as an effective method of signal analysis and non- stationary time series. In this study, by a combination of two instruments of fuzzy neural network and wavelet theory, a powerful model was created which is used to predict the qualitative parameters of sulfate in Hydrometric station of Chamriz (located on the Kor River). In general, in the wavelet-ANFIS model, results using Gaussian membership function leads to better results than the triangular membership function. It was also shown that by increasing the number of membership function, the accuracy of simulation results increases

concentration of sulfate, hybrid model of wavelet-ANFIS, Chamriz Station, Kor River