Fast Approach for Battery Impedance Identification Using Pseudo-Random Sequence Signals

Abstract

Online measurements of the battery impedance provide valuable information on the battery state-of-charge and state-of-health, which can be utilized for improving the safety and the performance of the associated system. The electrochemical-impedance spectroscopy (EIS) is widely used for battery impedance measurements, but it is not the most applicable solution for online measurements due to its slowness and complexity. These drawbacks can be improved using broadband signals, such as pseudorandom sequences (PRS), which are fast and easily implementable. However, the nonlinear behavior of batteries have a significant effect on the impedance measurements and the selection of the PRS signal. Majority of the PRS signals are applicable for measurements of linear systems, but also signals for nonlinear system identification do exist. Moreover, the reduced accuracy and signal-to-noise ratio of the PRS signals compared to the EIS make the filtering of the results as well as the amplitude design important aspects. This paper demonstrates the use of two PRS signals, the pseudorandom binary sequence (PRBS), and a ternary sequence with better toleration to battery nonlinear effects, with comprehensive amplitude and filtering design for battery impedance measurements. It is shown that the ternary sequence provides accurate measurements and the effects of nonlinear dynamics of the battery impedance are reduced with respect to the PRBS measurements. The results are referenced and validated to practical EIS measurements in various operating conditions for lithium-iron-phosphate ( $\text{LiFePO}_\text{4}$ ) cell.