Dynamic Tracking of Power Demand for Integrated Fuel Cell Systems using Nonlinear Model Predictive Control

Abstract

Transient changes in the power demand of state-of-the-art fuel cell systems are compensated by a battery in order to operate the fuel cell system safely within its physical boundaries. More concretely, oxygen starvation in the fuel cell is conventionally prevented by directly controlling the oxygen excess ratio. However, this limits the transient response of the fuel cell and the system’s overall flexibility and efficiency. In order to overcome these limitations, we ascribe the task of the dynamic but safe response in a hybrid system to the fuel cell. For this purpose, we present a nonlinear model predictive control approach which is able to realize efficient transient power tracking, while considering the oxygen excess ratio explicitly as a boundary. We address the control challenges of a nonlinear, coupled, and bounded system with an adequate control design using a real-time capable nonlinear controller model. The controller is validated as proof of concept in simulation with a detailed dynamic plant model. Our contribution realizes a collaborative power setting by fuel cell and compressor. Moreover, system efficiency both in stationary and in transient operation is achieved, while preventing oxygen starvation as well as compressor surge and choke throughout the entire operation.