Bilayered Real-Time Energy Management Strategy for Hybrid Power Systems in Hydrogen Fuel Cell Vessels
Yuji Zeng, Qinjin Zhang, Shi You, Yancheng Liu, Herbert Ho‐Ching Iu, Haohao Guo, Siyuan Liu
Abstract
Existing energy management strategies (EMSs) for hybrid power systems (HPSs) in hydrogen fuel cell vessels (FCVs) are not applicable to scenarios with multiple hydrogen fuel cells (FCs) and lithium batteries (LBs) in parallel and are difficult to achieve real-time control and optimization for multiple objectives. In this article, a bilayer real-time energy management strategy (BLRT-EMS) is proposed. Compared with existing EMSs, the proposed BLRT-EMS implements different control/optimization objectives distributed in the execution layer EMS (EL-EMS) and the decision layer EMS (DL-EMS), which can significantly reduce bus voltage fluctuations, decrease hydrogen consumptions, improve the system efficiency, and have potential for engineering applications. In the first EL-EMS, a decentralized optimal power allocation strategy is proposed, which allows each FC system (FCS) to allocate the output power ratio according to their generation costs, ensuring consistent performance of multiple FCSs (MFCSs) under long-term operating conditions, and thus delaying the degradation rate of FCs. In the second EL-EMS, a distributed cooperative control strategy is proposed to achieve dynamic state-of-charge (SoC) equalization, proportional output power allocation, and accurate bus voltage restoration among multiple battery storage systems (MBSSs) to extend the service life of batteries. In the DL-EMS, an energy coordination optimization strategy between MFCS and MBSS is proposed to achieve hydrogen consumption reduction and system efficiency improvement, thus enhancing the endurance performance of FCV. Finally, the test results based on the StarSim experimental platform show that the proposed BLRT-EMS has faster SoC convergence speed, smaller bus voltage deviation, lower hydrogen consumption, higher system efficiency, and lower operation stress than the state-of-the-art methods.