A New Noise Reduction Concept for the FFT EIS Method Using the Ergodic Hypothesis for Fuel Cell Measurement

Abstract

I propose a new noise reduction concept for FFT (Fast Fourier transform) Electrochemical Impedance Spectroscopy (EIS) using the ergodic hypothesis. Instead of a conventional time-averaged impedance, an ensemble average of impedances with slightly different frequencies (quasi-ensemble-median value) is defined as an experimentally obtained appropriate impedance. This approximation is acceptable in some conditions and will contribute to faster measurement. By applying this method to a hydrogen fuel cell, a high-accuracy measurement of fuel cell impedance between approx. 0.38 Hz and 10 kHz was performed within three seconds. The results from FRA and the proposed FFT method in the steady state were almost the same. The averaged percentage error throughout the whole of the measured frequencies was around ±1%, and the maximum difference was around −6% at approx. 1 Hz.