Low-carbon Economic Scheduling of Hydrogen Integrated Energy System Based on Optimistic Actor-critic Deep Reinforcement Learning
Under the background of"dual carbon",the hydrogen integrated energy system with the hydrogen energy as the energy carrier is an important support for the low-carbon transformation of China's energy industry.To ensure the supply and efficient utilization of the hydrogen energy in the hydrogen integrated energy system,this article proposes an operational mode of the integrated energy system which makes hydrogen energy from electricity and gas,which realizes the comprehensive utilization of the hydrogen energy.Based on the carbon capture devices and the integrated demand response,a carbon reduction mechanism with the complementarity between energy sources and loads is established to fully tap the potential of carbon reduction in the system,further improving the consumption rate of renewable energy and the low-carbon level of the system.In addition,in order to achieve a rapid response to the random fluctuations of the energy source and load in the integrated energy system containing hydrogen,an optimistic action-critic deep reinforcement learning method is proposed for the offline training and the online optimization of the scheduling model in the system,which efficiently achieves the low-carbon economic online optimization scheduling of the integrated energy system containing hydrogen energy.Finally,the superiority of the proposed method is verified through a responding case.
hydrogen integrated energy systemcomplementary carbon reductionlow-carbon economic schedulingdeep reinforcement learningcomprehensive utilization of hydrogen