Model Predictive Control of hybrid fuel cell/battery/supercapacitor power sources

Abstract

The use of fuel cell as future energy for vehicle application is very promising due to its environmentally friendly, efficient, and flexible properties. When applied to the vehicle, one drawback of fuel cells is that it is unable to supply sudden load changes since the dynamics of fuel cells is slow. However, this could be compensated by adding energy storage systems (batteries and/or supercapacitor). To supply the load, fuel cell and energy storage system will be controlled by a DC-DC converter. The constancy of DC output voltage (DC bus) in hybrid vehicle is very important even with large and rapid load changes. In addition, the increase in electrical current in fuel cells and batteries must be limited to extend their lifetime. In this study a control system was designed and implemented to regulate fuel cell as a primary energy source as well as batteries and supercapacitor as energy storage systems in order to obtain a constant DC output voltage and to limit current slope of fuel cell and batteries. There were three DC-DC converters deployed to regulate the output voltage of the three energy sources namely the boost converter, used to adjust the voltage/current output of the fuel cell, and two bidirectional converters, used to adjust the voltage/current output of energy storage systems (batteries and supercapacitor). To determine the reference current of each converter Model Predictive Control (MPC) was employed, and hysteresis control was functioned to track the reference current. MPC and hysteresis control were implemented on a dSPACE DS1104 Controller Board. For the experiment, small energy sources were operated, which were 50 W 10 A fuel cell, 6 V 4.5 Ah battery, and 7.5 V 120 F supercapacitor. Simulation and experiment results showed that the MPC can be designed to limit the current slope in fuel cells and batteries and to keep the bus voltage within the reference value.