Prediction of Wind Speed and Direction using Encoding - forecasting Network with Convolutional Long Short-term Memory

This paper presents the prediction system of wind speed and direction one hour ahead using encoding-forecasting network with convolutional long short-term memory (ConvLSTM). The input of prediction system is wind speed and direction which are represented as image data on the 2D coordinate and provided by Automated Meteorological Data Acquisition System (AMeDAS) in Japan. Performances of the proposed prediction system are evaluated based on root mean square error (RMSE) between observed and predicted value. The goal of the proposed prediction system is to improve prediction accuracy and it is confirmed by comparing the result of the prediction system of four seasons.