Battery Thermal Control: A Comparison between Predictive and Reactive Controllers

With the increase in hybrid and electric powertrains being developed, many concerns arise about the energy storage systems in all those vehicles. This unit supplies energy to every part, including its cooling system, so it becomes imperative that the BTMS balances the temperature and the energy spent on controlling it. This paper compares two fundamentally different control methods in four different test scenarios that simulate real situations faced in daily usage. The model is built digitally based on real NMC 21700 cells on Simcenter Amesim and then exported as an FMU file to MATLAB Simulink. The controllers were then created with the identified system and tuned to the FMU responses. The results indicate that the MPC can compensate for disturbances and act quickly on them, while the reactive nature of the PID takes longer to come into effect. However, the simulation with the MPC took much longer than the simpler PID, which can impact real-time situations, and the aggressive resulting nature of the predictive controller can lead to a higher cooling energy cost. Both had an acceptable response during all test cases and maintained the temperature despite the irregular heat generation profile in the cells. Future work will improve tuning and validate the controllers in real tests via hardware-in-the-loop methodology.