(363v) Online Impedance Analysis Using Chirp Signals in Linear and Non-Linear Systems

Frequency response analysis (FRA) for linear time invariant systems is well established and is used to characterise a system with the help of Nyquist, Bode and Nichols plots [1,2]. Electrochemical impedance spectroscopy (EIS) is nothing but FRA where impedance plot is the Nyquist plot generated using current as input and voltage as output. EIS is mainly used for diagnostics of various electrochemical systems. In case of batteries, it is used to gather information on state-of-charge & state-of-health of the battery [3]. The time required to carry out EIS is large, which prohibits its use for online rapid impedance measurement [4]. The intrinsic characteristic of instantaneous frequency for chirp input signal can be leveraged to mitigate the problem of sluggish nature of conventional EIS. For example, in the prior work of corresponding authors [5.6], it was shown that the chirp-based impedance for linear RLC circuits can be obtained in a few seconds as compared to hours using a conventional EIS. The generated impedance plots mirrored the plots generated through conventional EIS.

In this work, a mathematical foundation for the time-frequency equivalence of chirp signals is established for linear systems. Moreover, the chirp-based FRA is extended to non-linear systems using multiple chirp signals. The system is excited using multiple input chirp signals with the same frequency information but distinct initial phases. The corresponding output responses are modelled as chirp signals with a time-varying mean, time-varying amplitude ratio and time-varying phase lag. The varying mean enabled us to perform analysis on transient profile of the response along with steady-state response. Hence, we were able to capture the frequency response utilising entire output of chirp signals instead of waiting for the transients to disappear. This study is extended to nonlinear Li-ion battery systems to characterise the EIS plot. Simulations in MATLAB-Simulink environment were used to validate the technique. This proposed technique has the potential to be implemented commercially for online health monitoring of the electric vehicle batteries.

References:

1 Nyquist, H. Bell Syst. Tech. J. 11, 126–147 (1932)

2 Pintelon, R. et al. System Identification: A Frequency Domain Approach, Second Edition (Wiley, 2012)

3 Bonanos, N. et al. in Impedance Spectroscopy: Theory, Experiment, and Applications, Second Edition 205–537 (John Wiley & Sons, Inc., 2005)

4 Fasmin, F. et al. J. Electrochem. Soc. 164, H443–H455 (2017)

5 Bullecks, B. et al. Comput. Chem. Eng. 106, 421–436 (2017)

6 Suresh, R. et al. Int. J. Hydrogen Energy 45, 10536–10548 (2020)