Wind Speed and Power Prediction Using Artificial Neural Networks

Short-term wind prediction over different time steps is vital for wind farms in operation for various applications. Considering the complexity of atmospheric processes governing wind, time series models are preferred over physical models for wind prediction. Artificial neural networks (ANNs), which perform a non-linear mapping between inputs and outputs, provide an alternative approach for wind prediction. Two popular ANN types, Multi layer feed forward networks and Elman networks are developed for simulating wind speed and power for a typical Indian wind farm for different prediction lengths. The performance of the models is compared based on the RMS errors of the wind speed and power.