Digital Mirror Modeling Technique for Cell-Level Energy Storage Based on Two-Stage Dynamic Modeling

This paper proposes a digital mirror modeling technique for cell-level energy storage based on two-stage dynamic modeling. In the first stage, an autoregressive model of the lithiumion battery is established using an equivalent circuit model to form a real-time estimation of the equivalent physical parameters of the lithium-ion battery, and the accuracy of the model is verified through experiments. In the second stage, the historical data and model parameters of the battery are traversed and fitted over a complete State of Charge(SOC) cycle within several days/weeks, and the functional relationship between the physical parameters of the battery and the SOC is established and dynamically updated, so that accurate estimation and mirror simulation of the battery under different SOC states can be realized. By inputting the battery running currents into the mirror model, the error between the mirror model and the measured SOC is less than 3%.