Daily Average Wind Energy Forecasting Using Artificial Neural Networks

Abstract

This paper presents a proposal related to the wind power forecasting using artificial neural networks. An architecture based on feedforward artificial neural network is developed. The proposed solution is investigated in order to establish the optimum configuration and the performances in case of the South-East part of the Europe conditions. Conclusions and recommendations are emphasized based on the comparative study of the solution using solely long term historical wind power energy production and enhanced estimation performed with meteorological data related to the wind behavior over the studied period.