Thermal Management-Oriented Multivariable Robust Control of a kW-Scale Solid Oxide Fuel Cell Stand-Alone System

The aim of this paper is to develop an effective control system for a kW-scale solid oxide fuel cell (SOFC) stand-alone system based on our previous studies, where the dynamic model and the steady-state analysis and optimization of the system have been completed. The open-loop dynamic responses of the system are first conducted to gain insight on the system dynamics, and the results demonstrate that the system is a complicated multivariable system with strong coupling and nonlinear characteristics. Considering the complicated dynamics, a multivariable robust proportional-integral derivative (PID) control system is proposed with a multiloop feedforward/feedback control structure. The feedforward controllers are designed using the stack current based on the steady-state optimization results determined in the previous studies, and the feedback controllers are implemented by developing a single neuron adaptive PID algorithm. The adaptive PID combines the advantages of robust control and PID control, which can automatically adjust control parameters when system encounters uncertainties and disturbances, and is very easy for engineering practices because of its simple structure. Test results demonstrate that the multivariable robust PID control system has well robustness and stability for the SOFC system when operating point changes within the full operating range.