Time-domain system identification of Li-ion batteries from non-zero initial conditions
Résumé
Electrochemical Impedance Spectroscopy (EIS) allows characterizing electrochemical behavior of Lithium-ion batteries. However, it requires the battery to be at a relaxed state which is quite time-consuming during data acquisition process at different State-Of-Charge (SOC) levels. A time relaxation is usually observed of about 1h at each SOC level. In the prospect to reduce the measurement time, this paper presents a time-domain identification method using non-zero initial conditions of a fractional-order equivalent circuit model (FO-ECM). A two-stage iterative algorithm is developed. It estimates at one stage the system free response, due to the past initialization, and the system forced response due to the input/output signal. It uses, at the second stage, an output error model to estimate model parameters. The algorithm is iterated until convergence. The proposed identification algorithm is applied to identify a Li-ion battery FO-ECM using experimental data.