Dynamic economic scheduling of wind-solar complementary power generation system with electric-hydrogen hybrid energy storage based on deep reinforcement learning

To address the problem of renewable energy fluctuations in wind-photovoltaic (PV) power system with an electrochemical-hydrogen hybrid energy storage system, a dynamic economic dispatch method considering the operational characteristics of devices is proposed. First, the complementary characteristics of electrical energy storage and hydrogen energy storage are fully considered, and a mathematical description of the dynamic economic dispatch problem of the wind-PV power system with the hybrid energy storage system is established. Then, the dispatch problem is formulated as a Markov decision process framework, defining the system's observation state, dispatch action, and reward function. Subsequently, the dynamic economic dispatch decision is made using a dual deep Q-learning algorithm. Finally, the effectiveness of the proposed method is verified through simulation examples.