Identification of fractional order model for a voltammetric E-tongue system

Abstract

A method to estimate the fractional order model of a voltammetric electronic tongue (E-tongue) system with five working electrodes has been presented in this paper. The method uses fractional order system (FOS) identification to obtain a fractional order transfer function (FOTF) from the process data itself. Model quality measures of the proposed FOTF have been compared with that of the integral order transfer function (IOTF) which is based upon the Randles model for an electrochemical cell. It is found that FOTF produces better model performance than that of IOTF. The parameter estimation of the FOTF has been performed using the E-tongue response for five tea samples with different working electrodes. The cross-validation of FOTF has been carried out by obtaining the impedance spectra of the working electrodes. The efficacy of discrimination for the identified fractional model has been assessed by studying the clusters of response data for various tea samples.