Enhance the performance of weather parameters in Short-Term Weather forecasting using ANFIS

Weather prediction is an ever challenging area of investigation for scientists. The Adaptive Neuro-Fuzzy Inference System (ANFIS) has been widely used for modeling different kinds of nonlinear systems including rainfall forecasting. Adaptive Neuro-Fuzzy Inference Systems (ANFIS) combines the capabilities of Artificial Neural Networks (ANN) and Fuzzy Inference Systems (FIS) to solve different kinds of problems, especially efficient in rainfall prediction. In this paper the application of artificial neural networks to predict the Weather of Delhi city has been proposed using knowledge base in the Neuro-Fuzzy Inference system. The weather parameters like minimum temperature, maximum temperature, relative humidity, sea level pressure, rain fall, wind speed, wind direction and sun shine etc. has been used for prediction. When performing weather predictive model the key criteria is always accuracy. We are trying to predict future weather condition based upon above parameters by Artificial Neural Network. The model performance is contrasted with multi layered perceptron network. The proposed network train with actual data of the five years (2008 to 2012) of Safdarjung station, New Delhi and tested which comes from meteorological department. MLP used with Fuzzy logic. KeywordsAdaptive Neuro-Fuzzy Inference systems (ANFIS), Weather prediction, RMSE, MSE, RMASE, Coefficient Correlation, SNR.