Digital Twin Modeling Using High-Fidelity Battery Models for State Estimation and Control

Lithium-ion batteries (LIBs) play a vital role in the advancement of electric vehicles and sustainable energy solutions. They are favored over other secondary energy storage systems due to their high energy density, long cycle life, high nominal voltage, and low self-discharge rate. However, the latency of its internal states makes it difficult to predict its performance and ensure it is being operated safely. Fortunately, battery management systems (BMS) can use battery models to predict the internal states of a battery. There is a constant trade-off between accuracy and computational cost when it comes to battery models with only a handful being able to meet the constraints of a BMS. The following paper will showcase a Digital Twin framework that captures the accuracy of high-fidelity electrochemical models while meeting the computational constraints imposed by the BMS. The proposed framework will show that a high-fidelity model can be used to predict slower dynamics such as the state of health (SOH) and more dynamic states such as voltage, temperature, and state of charge (SOC) can be accurately predicted using a lower-fidelity model in Real-Time.