Multi-timescale Optimal Scheduling of an Electricity-Hydrogen Integrated Energy System Based on Deep Deterministic Policy Gradient

The variability in the time scales of electric and hydrogen energy scheduling of an electricity-hydrogen integrated energy system (EH-IES) poses many challenges to the efficient and economic management. To this end, the paper presents a multi-timescale optimal scheduling approach of the EH-IES based on deep deterministic policy gradient (DDPG) algorithm. Firstly, the EH-IES model is introduced, focusing on the hydrogen storage model, the electricity storage model and gas turbines operation model. Secondly, the EH-IES two-layer optimal scheduling model with multi-timescale is established. The upper layer is the storage and usage scheduling of hydrogen energy for 7 days on the long time scales, and the lower layer is the charging and discharging scheduling of electric energy for 24 hours on the short time scales. Then the DDPG algorithm is adopted and the two-layer optimal scheduling model is solved to achieve the dynamic scheduling of the EH-IES. Lastly, analysis of a case is executed to confirm the effectiveness of the proposed method.