3Department of Civil Engineering, Tabriz University
Analyses and investigations on river flow behavior are major issues in design, operation and studies related to water engineering. Thus, recently the application of chaos theory and new techniques, such as chaos theory, has been considered in hydrology and water resources due to relevant innovations and ability. This paper compares the performance of chaos theory with Anfis model and discusses on application case in the context of different interpretations of chaotic behaviour in river flow time series. This study determines the daily flow properties of river Aharchai in during 19 years using the concepts of chaos theory and predicted flows. Reconstruction of state space time series using chaos theory, based on appropriate selection of delay time and embedding dimension. Average mutual Correlation dimension technique has been used for definition of fractal dimension and evaluation of chaos in time series.Results of Evaluationsshow the fractals dimension of 4 (chaotic low), with a time delay of 65 days and embedding dimension of 13 that can be used for the reconstruction of dynamic state space of river flow. Local prediction algorithm is used for prediction of the time series. The results represent acceptable precision and adequate theory of chaos in flow forecasting of Aharchai River.