Assessment of Different Deep Learning Models for Hardware Requirements and Pre-Processing Steps for Efficient Wind Power Forecasting

Wind power generating systems are increasing their presence in the whole energy fabric. However, intermittent energy production by wind power plants due to its environmental dependency poses a threat to the energy grid. To address this challenge of uncertainty, auxiliary energy storage units or accurate and reliable forecasting are two potential solutions. However, integration of energy storage units will increase the capital investment and in turn, will decrease the return on investment of wind power plants. So, an accurate forecasting mechanism is the most promising way to address the challenges of uncertainty. In recent years, lots of research works have proposed different Machine Learning (ML) or Deep Learning (DL) approaches for accurate wind power forecasting. In this proposed work different DL models are analyzed for wind power forecasting with different pre-processing methods. So, the major contribution of the work lies in the finding an optimal deep-learning model and pre-processing strategy for improving wind power prediction accuracy.