Introduction: Proton exchange membrane fuel cells (PEMFCs) are highly efficient and environmentally friendly energy conversion devices. Their modular multi-stack FC (MSFC) provides enhanced operational stability and power output. However, during online electrochemical impedance spectroscopy (EIS) testing, multi-frequency sinusoidal current disturbances injected by the DC/DC converter can cause periodic fluctuations in the common DC bus voltage, threatening system stability and measurement accuracy. Existing approaches, such as hardware filtering or improved topology, suffer from high cost and complexity. While the traditional extended state observer (ESO) can estimate the disturbance, it struggles to suppress the fluctuations during multi-frequency co-detection. Methods: This paper proposes an enhanced ESO design, namely the multi-resonant ESO (MRESO), for MSFC hybrid systems. By embedding multiple resonant units, the MRESO accurately tracks the EIS disturbance frequency and combines feedforward compensation with battery-assisted control to achieve robust bus voltage stability. The system modeling is based on the principle of energy conservation, and the control scheme adopts a dual-loop structure consisting of an inner current loop and an outer voltage loop. The voltage loop integrates the MRESO to improve interference rejection capability. Results: Simulation results show that under load current disturbances and EIS measurement interference, MRESO can reduce voltage fluctuations by over 70% and achieve stability within 0.08 s, outperforming both PI control and traditional ESO, validating its effectiveness in suppressing multi-frequency disturbances. Discussion: This study provides a feasible voltage stabilization solution for online EIS monitoring of MSFC, contributing to improve the power quality and reliability of the hybrid system.

Precise prediction of voltage degradation is critical for the prognostics and health management of proton exchange membrane fuel cells (PEMFCs). The performance degradation of PEMFCs is governed by a complex coupling of multiple physicochemical mechanisms, including catalyst layer and proton exchange membrane degradation. Crucially, these internal degradation processes evolve across highly heterogeneous time scales, ranging from transient high-frequency fluctuations to low-frequency and long-term irreversible performance fade. Conventional predictive models, which typically rely on single-scale architectures or fixed receptive fields, are inherently ill-equipped to simultaneously decouple and capture these cross-scale temporal dynamics. To tackle this challenge, this paper innovatively proposes a multi-scale deep learning framework that integrates a multi-scale degradation trend perception module, a long short-term memory (LSTM)-based encoder–decoder architecture, and a multi-head attention mechanism. One-dimensional convolutional layers with different kernel sizes are employed to simultaneously extract local temporal features at multiple granularities, followed by the LSTM encoder–decoder to model long-range temporal dependencies, while the cross-attention mechanism dynamically allocates attention across the encoded context at each autoregressive decoding step. Experimental outcomes indicate that the proposed model realizes excellent predictive accuracy across five evaluation indices in comparison with standard baselines. In particular, the mean absolute percentage error (MAPE) reaches 1.6696%, and the maximum absolute percentage error (Max-APE) is strictly bounded within 5%, substantiating the reliability of the proposed framework for high precision and long-horizon health prognostics for PEMFCs.

Efficient thermal management is critical for proton exchange membrane fuel cells (PEMFCs). This study develops a bio-inspired active liquid cooling architecture to overcome the limitations of conventional flow channels, where convective heat transfer augmentation significantly increases pressure drop. The proposed configuration adheres to constructal theory and Murray’s law. A three-dimensional conjugate heat transfer model was formulated to evaluate the thermo-hydrodynamic performance against parallel and serpentine flow channels. Under identical conditions, the proposed configuration exhibits superior thermal uniformity and hydrodynamic behavior, alongside minimized parasitic pumping power. At an inlet Reynolds number (Re) of 400, this configuration stabilizes the average bipolar plate temperature at 349.63 K. It reduces the index of uniform temperature (IUT) to 1.49 K, representing a 52.8% thermal uniformity improvement over the parallel flow channel. Furthermore, at an inlet Re of 600, the overall pressure drop is restricted to 68.44 Pa, reducing the single-plate parasitic pumping power to 1.27 × 10⁻⁴ W, which represents reductions of 93.5% and 12.5% relative to the serpentine and parallel flow channels, respectively. This study provides an alternative architectural scheme for the design of active liquid cooling flow channels in PEMFCs.