Variable Nozzle Turbine‐Based Air Compressor Energy Recovery From High‐Power Fuel Cell System Exhaust via Adaptive Model Predictive Control

Abstract This study proposes a variable nozzle turbine (VNT)‐based air compressor system that recovers energy from high‐power vehicular fuel cell exhaust to reduce parasitic power and enhance efficiency. The system utilizes exhaust gas kinetic energy to drive a turbine connected to a high‐speed compressor motor via the main shaft. By dynamically adjusting the nozzle opening degree (NOD), the turbine optimizes exhaust flow area and backpressure, maximizing energy recovery power across operating conditions. Simultaneously, regulating the oxygen excess ratio (OER) ensures optimal cathode oxygen supply for the proton exchange membrane fuel cell (PEMFC). A control‐oriented model capturing VNT‐based air compressor and PEMFC coupling dynamics is developed and validated. Multi‐start sequential quadratic programming identifies optimal OER trajectories that minimize compressor motor power while maintaining electrochemical performance. To track the target, a linear parameter variation adaptive model predictive control is implemented. Relationships between NOD, OER, and PEMFC power characteristics are embedded in a control‐oriented model validated through map testing. The proposed strategy demonstrates superior OER tracking under dynamic loads like the China Heavy‐duty Truck Cycle. The integrated system is 3.5788% higher than fixed‐OER strategies with delivering 20.5028 kWh net energy output, it demonstrating effective coordination of energy recovery and oxygen regulation.