基于混合规则与优化的新型多堆集成氢储能系统能量管理策略

Abstract To improve the performance of off-grid energy systems , based on a novel multi-stack integrated hydrogen energy storage system, a full life cycle energy management strategy (EMS) with hybrid rules and optimization is proposed. Such a system consists of four PEMECs, four PEMFCs, and a battery pack. The proposed EMS includes online parameter estimation, rule-based primary allocation, and optimization-based secondary allocation. The power is firstly allocated among PEMEC/PEMFC stacks according to a daisy-chained rule based on high efficiency range and multi-stack voltage degradation consistency. The remaining fluctuating power is then preferentially absorbed by the battery using a Q-learning reinforcement learning algorithm. Under stochastic wind power, photovoltaic , and regular load, a comparative analysis with three common EMS validates the feasibility of the proposed EMS. Compared to Q-DC and SOC-DC, the voltage degradation of the multi-stack PEMFCs is reduced by 13.9 %, and the lifetime is improved by 17.2 % and 15.7 %, saving 1315 kg and 1190 kg of hydrogen, respectively. Compared to SOC-PDC and SOC-DC, the lifetime of the battery is improved by 18.8 % and 20.8 %, respectively. The annualized cost of system based on the proposed EMS is the smallest, with reductions of 3 %, 7.4 %, and 9.2 %, respectively.