Fuzzy model identification based on cluster estimation for reservoir inflow forecasting.

The article discusses a study about the impact of choice of clustering algorithm on the overall performance of a fuzzy-based hydrologic model. An overview of the Takagi-Sugeno (TS) fuzzy model is presented, along with a brief discussion about identification of the TS fuzzy model including the optimal number of fuzzy partitions. The study is based on a research about developing a TS fuzzy model for reservoir inflow forecasting in the Narmada basin, India. Key findings suggest that the choice of the clustering algorithm may not have a significant impact on the model performance if the forecast needed at 1 hour in advance.