For accurate and efficient use of wind power, it is important to know the statistical characteristics, availability, diurnal variation, and prediction of wind speed. Prediction of wind power permits the scheduling of the connection or the disconnection of wind turbines to achieve optimal operating costs. In this paper, a simple and accurate method for predicting wind speed based on weather-sensitive data is presented. The proposed wind speed prediction system is cost-effective and only needs wind speed data at 40 m and weather data to forecast wind speeds at 50 m and 60 m for the current and next months. Hellman coefficients are first estimated by using a feed-forward backpropagation neural network and wind speeds at different heights are predicted. The autoregressive moving average algorithm is used for forecasting the short-term wind speed and is compared to in situ measurements. The predicted results are then compared to a powerful estimation algorithm known as the Mycielski algorithm.

For accurate and efficient use of wind power, it is important to know the statistical characteristics, availability, diurnal variation, and prediction of wind speed. Prediction of wind power permits the scheduling of the connection or the disconnection of wind turbines to achieve optimal operating costs. In this paper, a simple and accurate method for predicting wind speed based on weather-sensitive data is presented. The proposed wind speed prediction system is cost-effective and only needs wind speed data at 40 m and weather data to forecast wind speeds at 50 m and 60 m for the current and next months. Hellman coefficients are first estimated by using a feed-forward backpropagation neural network and wind speeds at different heights are predicted. The autoregressive moving average algorithm is used for forecasting the short-term wind speed and is compared to in situ measurements. The predicted results are then compared to a powerful estimation algorithm known as the Mycielski algorithm.