Zeros-MTL-CLDBN: A Robust Multitask Deep-Learning Framework via Deep-Belief Network with Correntropy Loss for Ultrashort-Term Multistep Wind Power Forecasting

Abstract

AbstractAccurate ultrashort-term multistep wind power forecasting (WPF) remains a critical challenge in renewable energy systems due to the inherent stochasticity, non-Gaussian fluctuations, and volatility of wind energy. To address this challenge, this ...