Affiliation:
1. Tongji University, School of Automotive Studies
2. Tongji University
3. Shanghai Zhongqiao Vocational and Technical University
Abstract
<div class="section abstract"><div class="htmlview paragraph">The multi-stack fuel cell system (MFCS) has the advantages of higher efficiency,
stronger robustness and longer life, and could be widely used in high-power
application scenarios such as automobiles, airplanes, trains, and ships. The
appropriate air mass flow and air pressure have a crucial impact on the output
power performance indicators of the MFCS. Considering that the designed
integrated air supply system for the MFCS has significant gas supply hysteresis
and strong coupling between the inlet air mass flow and air pressure of each
stack, this paper identifies multiple steady-state operating points of the fuel
cell system to obtain corresponding linear predictive models and establishes
corresponding predictive control algorithms. The Model Predictive Control (MPC)
algorithms are switched in real-time based on the current load throughout the
entire C-WTVC (China World Transient Vehicle Cycle) working condition. The
simulation results show that the designed MPC algorithm can control all inlet
air flow and air pressure of the MFCS (20kW/70kW/120kW) within the error range
of ± 2% of the expected target values, which is significantly better than the
PID control algorithm.</div></div>