Abstract
Optimal temperature regulation is crucial for the output capacity and durability of proton exchange membrane fuel cells (PEMFCs) under dynamic load conditions. Conventional temperature control strategies separate offline optimal determination from online tracking. Due to the lack of real-time thermal dynamics feedback, these two-stage methods exhibit limited adaptability. This article proposes a hierarchical control strategy that integrates online temperature optimization and regulation. The upper level combines polarization curve-based optimal temperature calibration with a dynamic protection mechanism. Real-time stack humidity feedback enables online adjustment of operating temperature. The lower level employs an improved generalized data-enabled predictive control (GDeePC) algorithm to achieve fast, precise, and model-free temperature tracking. Two key enhancements are incorporated: filtering with adaptive forgetting factors for historical data utilization, and a generalized cost function with tracking error derivatives to enhance transient response. Finally, the hardware-in-loop experimental results demonstrate the proposed strategy’s superior performance compared to conventional methods, particularly in maintaining temperature stability during load transients while minimizing overshoot and steady-state error.
Funding Information
  • Zhejiang Provincial Natural Science Foundation of China (LZ23F030008)
  • National Natural Science Foundation of China (62203386)
  • State Key Laboratory of Industrial Control Technology, China (ICT2024A11)

This publication has 35 references indexed in Scilit: