Deep Reinforcement Learning Based Real-time AC Optimal Power Flow Considering Uncertainties

Modern power systems are experiencing larger fluctuations and more uncertainties caused by increased penetration of renewable energy sources (RESs) electronics equipment. Therefore, fast accurate corrective control actions in real time needed to ensure the system security economics. This paper presents a novel method derive real-time alternating current (AC) optimal flow (OPF) solutions considering including varying topology changes using state-of-the-art deep rein-forcement learning (DRL) algorithm, which can effectively assist grid operators making rapid effective decisions. The presented DRL-based approach first adopts super-vised-learning from generate good initial weights for neural networks, then proximal policy optimization (PPO) algorithm is applied train test artificial intelligence (AI) agents stable robust performance. An ancillary classifier designed identify feasibility AC OPF problem. Case studies conducted on Illi-nois 200-bus with wind generation variation N-1 validate effectiveness proposed demonstrate its great potential promoting sustainable integration into system.