Deep Learning-Based Adaptive Remedial Action Scheme with Security Margin for Renewable-Dominated Power Grids

The Remedial Action Scheme (RAS) is designed to take corrective actions after detecting predetermined conditions maintain system transient stability in large interconnected power grids. However, since RAS usually based on a few selected typical operating conditions, it not optimal that are considered the offline design, especially under frequently and dramatically varying due increasing integration of intermittent renewables. deep learning-based proposed enhance adaptivity conditions. During training, customized loss function developed penalize negative suggest with security margin avoid triggering under-frequency over-frequency relays. Simulation results reduced United States Western Interconnection model demonstrate learning–based can provide for unseen while maintaining sufficient margin.