Abstract
Measurements acquired on batteries in the form of time signals such as voltage-time and capacity-time to assess their cyclability performance can be supplemented by examining their frequency-domain response. This allows one to determine the global characteristics of the signals and the battery, but not the local behavior, which is very important for determining for example battery fading. In this study we examine the short-time Fourier transform for time-frequency deconstruction of galvanostatic charge/discharge signals of lithium-sulfur batteries, taken as an example. The results displayed in terms of spectrograms show how the frequency content of such signals (e.g. charge and voltage time series) evolve with the lifetime of the batteries allowing the detection of critical changes in the response that may lead to fading and eventually default.
