Towards practical impedance-based temperature estimation of Li-ion batteries: impedance measurements and feature extraction

Abstract

The temperature of Lithium-ion (Li-ion) batteries have been recently reported to be accessible without external sensor via battery impedance. For practical impedance-based temperature estimation, it is beneficial that the impedance is comprehensively studied to find out the frequency bands with the most characteristic dependency on the battery temperature to increase the reliability of the estimation in varying operating conditions. In addition, the applied impedance measurement method should be fast, accurate and easy to implement for practical applications. This paper shows and compares feature extraction methods for battery temperature estimation based on practical impedance measurements utilizing pseudo-random-sequence (PRS) signals. The proposed methods are demonstrated by experiments carried out on a nickel-manganese-cobalt (NMC) Li-ion cell in wide range of SOCs and temperatures.