Smart sensor systems using impedance spectroscopy

Abstract

Extracting relevant information from impedance spectra is a big challenge if we consider different mechanisms, including cable and contacting effects, material properties and system instability. A particularly important prerequisite is that these effects depend differently on the frequency and that some measures can be taken to eliminate effects of irrelevant mechanisms. The common way in signal processing is to model the impedance spectrum by means of equivalent circuits and to extract the quantity of interest using optimization techniques. The more effects and mechanisms are represented in an impedance spectrum, the more unknown parameters are needed for accurate modelling. The parameter extraction process becomes thereby difficult to solve and needs a lot of trials and a priori knowledge. It is therefore important that mechanisms with minor influence are eliminated or neglected. Other possibilities for signal processing provide mathematical data fusion methods. Statistical methods like principal component analysis can be used for extracting specific information from the measured data. The paper considers different applications (electrochemical, biomedical, moisture analysis, battery diagnosing) of electrical impedance spectroscopy.