3D commercial software with multiphysics modeling is able to provide detailed solutions on battery cells and power electronics cooling. However, it is usually time-consuming and not compatible with the models of battery management systems (BMS).
Alternatively, engineers were seeking to build a 0D/1D thermal model, for example a thermal network, for reducing computational overheads. Meanwhile, such a thermal network approach typically has case-by-case fidelity and requires a vast number of human labors on manual parameter tuning and optimization.
The Reduced Order Model (ROM) is a systematic methodology to build such a thermal model. It is realized by state space model, which is fitted to have the heat transfer function calculated from 3D results.
A ROM can be generated via a well-defined automatic process in ANSYS TwinBuilder with well-controlled accuracy. A ROM can then be exported into a functional mock-up unit (FMU) and shared or utilized by various systems accepting FMUs.
In this presentation, an electro-thermal coupled battery model for an A123 liquid-cooled 48V Li-on battery pack is developed. The electrical part in the coupled model uses the equivalent circuit model (ECM) approach due to its speed and accuracy. The thermal part in the coupled model uses the ROM approach.
Compared with the common thermal network approach, the ROM approach demonstrates higher level accuracy and convenience. The coupled model, as part of a BMS, is then simulated in Matlab Simulink.
The capability of the coupled model on developing derate function in battery state of power (SoP) algorithm is demonstrated. It is shown that current derate caused by battery overheating is accurately captured under a dynamic drive cycle current profile. Such a ROM-based electro-thermal coupled model proves to be a powerful tool for BMS development.