Design and experimental study of energy management strategies for fuel cell vehicles

For achieving the global “Dual Carbon” goals, the transportation sector urgently needs cleaner and more efficient powertrains. The proton exchange membrane fuel cell (PEMFC) technology, featuring zero tailpipe emissions, high energy density, and extended range, has become a key means for decarbonizing commercial vehicles. However, the slow dynamic response of PEMFC systems makes it difficult for a stand-alone fuel cell to withstand frequent start-stop cycles and rapid load fluctuations. Therefore, fuel cell–battery hybrid architectures are widely adopted, placing higher requirements on energy-management strategies (EMSs) to efficiently coordinate multiple energy sources. This study addresses the EMS design problem for fuel cell commercial vehicles and experimentally compares two representative strategies from an engineering perspective to provide a basis for system optimization.

A hardware-in-the-loop bench platform for a fuel cell-battery hybrid electric vehicle was established. The platform integrated a PEMFC system, a lithium iron phosphate battery pack, and a drive motor. Further, a dynamic load simulation system was used to reproduce real-world driving resistances. A self-developed LabVIEW-based software platform, combined with controller area network bus communication, enabled real-time driving-cycle replay, data acquisition, and EMS verification. Using this platform, two EMSs were implemented: a deterministic rule-based strategy and a fuzzy logic control (FLC)-based strategy. The rule-based strategy adopted a master-slave architecture, in which the fuel cell served as the main power source, and power was allocated according to total vehicle demand and battery state-of-charge (SOC) thresholds. The FLC-based strategy used vehicle power demand and the SOC as inputs and outputted the fuel cell power, adjusting power distribution based on predefined fuzzy rules.

Bench tests under the same driving cycle revealed distinct characteristics of the two strategies. Both strategies achieved similar peak and average motor power; however, the FLC-based strategy exhibited slightly larger power fluctuations and faster tracking of load changes, indicating better dynamic adaptability and smoother acceleration and deceleration behavior. In the FLC-based strategy, the fuel cell operated at a lower average power with more frequent start–stop cycles and the battery contributed more to power regulation. In contrast, in the rule-based strategy, the fuel cell output was more continuous and often close to its rated power. Further, the battery SOC was maintained within a narrow band, benefiting battery life. Hydrogen consumption analysis showed that although the FLC reduced instantaneous fuel cell usage, its higher reliance on battery charge-discharge cycling led to higher equivalent hydrogen consumption. In contrast, the rule-based strategy offered shorter energy paths and better overall fuel economy. Quantitatively, the rule-based strategy achieved an actual and equivalent hydrogen consumption of 568.13 and 628.84 g, respectively. Meanwhile, the FLC-based strategy achieved 441.41 and 726.40 g, respectively.

The two EMSs present complementary advantages. The FLC-based strategy is more suitable for urban scenarios with frequent start–stop cycles and rapidly varying loads, where dynamic response and driving smoothness are prioritized. Conversely, the deterministic rule-based strategy is more appropriate for stable operating conditions that prioritize driving range and hydrogen economy, owing to its efficient and stable power-supply characteristics. Future work will focus on adaptive or hybrid EMS frameworks that combine rule-based logic with data-driven or intelligent optimization methods to enhance multiscenario applicability and overall energy-management performance of fuel cell-battery hybrid commercial vehicles.