An Artificial Neural Network Model for Wind Energy Estimation
IEEE SoutheastCon 2015 Conference
Wind energy resources are ideally suited for distributed generation systems to provide electricity for residential use. This paper proposes a novel method for wind energy estimation in the state of Georgia. This method is based on Artificial Neural Network (ANN) using real data obtained from several weather station sites around the state. The proposed ANN model was trained and then tested using a local station located in Savannah. The ANN inputs are elevation, latitude, longitude, day, temperatures (min/max), and the output is the daily wind speed. The model was efficiently implemented in Simulink environment using closed-form algebraic equations which eliminated the need for repeated training. The ANN model was formulated with suitable numbers of layers/neurons which was trained and tested with excellent regression constant. Furthermore, the ANN model has the ability to interpolate between learning curves to generate wind speed estimates for different locations. It is anticipated that this model will be able to successfully select sites for wind turbine installations for residential applications in the state of Georgia.
El Shahat, Adel, Rami Haddad, Youakim Kalaani.
"An Artificial Neural Network Model for Wind Energy Estimation."
IEEE SoutheastCon 2015 Conference Fort Lauderdale, Florida.