Indian Weather Forecasting using ANFIS and ARIMA based Interval Type-2 Fuzzy Logic Model

This paper presents a comprehensive study of ANFIS+ARIMA+IT2FLS models for forecasting the weather of Raipur, Chhattisgarh, India. For developing the models, ten year data (2000-2009) comprising daily average temperature (dry-wet), air pressure, and wind-speed etc. have been used. Adaptive Network Based Fuzzy Inference System (ANFIS) and Auto Regressive Moving Average (ARIMA) models based on Interval Type2 Fuzzy logic System (IT2FLS) have been applied. To ensure the effectiveness of ARIMA+IT2FLS and ANFIS techniques, different models employing a different training and test data set have been tested. The criteria of performance evaluation are calculated for estimating and comparing the performances of ARIMA+IT2FLS and ANFIS models. The performance comparisons of ANFIS and ARIMA+IT2FLS models due to MAE (Moving Average Error), RMSER (Root-Mean-Square error) criteria, indicate that ANFIS yields better results. Interval Type-2 Fuzzy time series models have been proposed for forecasting temperature, pressure, wind speed and other weather parameters. In this paper a hybrid fuzzy time series model is proposed that will develop Interval type 2 fuzzy models based on ARIMA. The proposed model will use ARIMA to select appropriate coefficients from the observed dataset. IT2-FLS is utilized here for handling the uncertainty in the time series data so that it may yield a more accurate forecasting result.