Economic Model Predictive Control for PEM Electrolyzers With DC/DC Interface Supporting Flexible Power Point Tracking

Given their rapid dynamic response and broad power-tracking capability, proton exchange membrane electrolyzers (PEMELs) are increasingly utilized as loads for flexible power point tracking (FPPT), facilitating the large-scale integration of intermittent renewable energy sources into power systems. However, achieving coordinated control of FPPT and thermal regulation in the PEMEL system is challenging due to its inherent multi-timescale dynamics. To address this issue, an advanced two-stage economic model predictive control (EMPC) scheme is proposed for PEMEL systems to enhance their dynamic economic performance, including Stage-1 and Stage-2 EMPCs. Based on a fourth-order thermal dynamic model that incorporates the stack's voltage-current characteristics, Stage-1 EMPC integrates real-time economic optimization and thermal regulation into one optimal control framework, achieving FPPT and reduced valve losses. Stage-2 EMPC, based on a DC/DC converter model, directly controls the switching signal to accurately track the current set-point while minimizing switching losses. Simulation results under varying power-supply conditions demonstrate that the proposed two-stage EMPC ensures effective hydrogen production while enhancing dynamic economic performance, with robustness guaranteed.