Resilient Distributed Generation with Hydrogen Fuel Cells: A Hierarchical Approach

Hydrogen fuel cells (FCs) have been widely regarded as a promising solution for zero carbon distributed power generation. However, despite the demand for network flexibility and plug-and-play capabilities, existing research on distributed generation with fuel cells has been largely limited to the centralized framework. This paper develops a distributed hierarchical controller that enables networked hydrogen fuel cells to reach exact averaging of the sum of their load demand, even in the presence of varying topology and discontinuous communication. In the upper (cyber) layer, a novel distributed event-triggered average consensus observer is proposed. Different from existing dynamic averaging observers, the network topology is allowed to be switching and jointly connected, the load demand signal is allowed to be uncertain, and the event-triggering protocol can be designed locally. In the lower (physical) layer, a multi-physical control-oriented fuel cell model that integrates the electrical and physical dynamics is established, and an adaptive controller is designed to simultaneously guarantee asymptotic tracking and L_∞ tracking performance. Moreover, it is shown that with the proposed approach, load sharing performance can be sustained even when new agents join or leave the network permanently. Finally, the performance of the proposed controller is validated on an experimental platform consisting of four networked hydrogen fuel cells. To the best of the authors' knowledge, this is the first physical validation of multi-fuel cell generation with dynamic topology. © 2025 IEEE.