Offline order recognition for state estimation of Lithium-ion battery using fractional order model

The orders of the fractional order model (FOM) for lithium-ion battery (LIB) inherit profound electrochemical significance, and their casual assignment and acquisition could lead to physically unreasonable value and hence the numerical instability, which can be fatal especially for on-vehicle applications. To obtain the order accurately, approaches based on the electrochemical impedance spectroscopy (EIS) analysis and the distribution of relaxation time (DRT) transformation are proposed in this paper. Correction of the raw EIS data eliminates the misinterpretation of the order due to parasitic inductance. DRT peak manipulation further facilitates the alteration and isolation of electrochemical processes, allowing concentration on the interested kinetics. The treatments along with their mathematical basis are elaborated, outlining a feasible scheme for the implementation of FOM. Based on the FOMs inspired by the EIS landscape, validity of the ideas is verified under different conditions. Satisfactory estimations of the state of charge (SOC) are achieved with the profiles of fractional order, and the potential diagnosis for the state of health (SOH) is also outlined. This paper aims to provide an engineering-friendly implementation of FOM for the battery management on electrified vehicles.