Robust Predictive Battery Thermal Management Strategy for Connected and Automated Hybrid Electric Vehicles Based on Thermoelectric Parameter Uncertainty

The connected and automated hybrid electric vehicles (CAHEVs) have the potential to improve safety and mitigate traffic congestions. A crucial problem of the CAHEVs is that the lithium-ion batteries are highly temperature sensitive, whose output power is severely limited at low temperatures. Moreover, the cells may be premature aging at high operating temperatures and even result in an explosion, seriously threatening the human safety. Consequently, a practical and energy-efficient battery thermal management (BTM) strategy is required with the minimum possible cooling/heating energy consumption. To achieve the multiple objectives, a finite-set based model predictive control (FSMPC) strategy is presented for the BTM in CAHEVs. Since the thermoelectric model of the BTM system is highly nonlinear and time variant, an extended state observer is implemented for accurate system state estimation and prediction so that the parameter uncertainties inherent in the system can be compensated. The hardware-in-the-loop validation of the proposed strategy is conducted on the urban dynamometer drive schedule-based on a Toyota Prius plug-in HEV model. The results demonstrate that the proposed BTM strategy can maintain the battery pack in the CAHEV operating at the optimum temperature and save 30% BTM energy compared to the conventional method even under 50% parameter uncertainty.