Rapid Broadband Impedance Acquisition of Lithium-Ion Batteries Based on Measurement Evaluating and Impedance Filtering

Abstract

Battery impedance contains valuable information concerning state estimation and fault diagnosis, while online impedance measurements attract increasing research interest. The Pseudo Random Sequence (PRS) signal enables a fast and low-complexity impedance measurement for Lithium-ion (Li-ion) batteries yet with the drawback of an inadequate Signal-to-Noise Ratio (SNR) that easily leads to a biased result, while very few studies have existed to process the non-uniformly distributed impedance data with large biases. Motivated by this, this work proposes a novel Measurement Evaluating Procedure (MEP) and a 2-Dimensional Gaussian Filter (2-DGF) for enhancing the performance of the PRS method. The MEP weights the impedance measurement data based on the signal power spectrum and determines the Gaussian Probability Distributions (GPDs) of the impedance data to eliminate interference from the abnormal data. The 2-DGF is further carried out on a semi-logarithmic scale to process the non-uniform impedance measurements, which avoids constantly adjusting the filtering windows. Experimental studies validate the proposed method under the different States of Charges (SOCs) and ambient temperatures. The results present the good accuracy and robustness of the proposed method under different cases.