Virtual Approach for Control System Design: Integrated Simulation of Battery Cooling and Cabin Comfort Circuits to Develop a BEV Thermal Management Control Logic

摘要

The automotive industry is following the worldwide vehicle and powertrain electrification trend, which is expected to increasingly grow in next decades. Electrified vehicles represent the main challenge for OEMs and first-tier suppliers which are requested to operate in a relatively new field, facing additional constraints, issues and requirements. A key point for hybrid and electric vehicles design is the proper sizing and control strategy development of the cooling and heating system, which is not only required to maintain the battery within the optimal temperature operating range and guarantee appropriate cooling of the other electrical components, but also to ensure thermal comfort for the passengers. In fact, the battery cooling system must be strictly linked to the AC circuit to guarantee cooling capacity whenever the ambient conditions do not allow passive cooling with conventional radiators, and linked to High Voltage Heater loop to provide fast warm up of the battery when operating in cold climate, thus adding several degrees of complexity to the management and calibration of a HEV/BEV cooling system characterized by different interacting subsystems. In an attempt to address the difficulties arising in the development of cooling systems for electrified vehicles, this paper proposes an effective approach based on CAE methods to speed up and support the system design and control development. A fully-physical 1D model of the battery cooling circuit and HVAC loop - representing a BEV system layout currently under study - is built in GT-SUITE and used as a plant for control strategies development. A robust and comprehensive control logic capable of managing the cooling and HVAC systems under different operating conditions is developed in MATLAB-Simulink and virtually validated by means of a coupled simulation with the GT-SUITE model. Following the normal product development process, the MATLAB-Simulink model could be eventually deployed in a real control unit to assess and furtherly validate the proposed strategy.