Multi-Agent System for Weather Forecasting

Weather prediction is a challenging task carried out by Meteorologist. Weather events impact a country’s social balance. This paper discusses a novel idea of using Multi Agent System (MAS) for Weather Forecasting. The proposed approach incorporates a deep neural network model within MAS with a hybrid ANN algorithm to recognize the static and dynamic weather conditions. The approach uses ensemble prediction to account for indeterminism in weather conditions. Predictions and alerts given by MAS can help government and local authorities to plan precautions in a timely manner. The paper discusses the implementation challenges and advantages of MAS Model compared to Numerical Weather Prediction Method