Modelling of direction-dependent dynamic processes: a comparison of Wiener models and neural networks

Abstract

The modelling of direction-dependent processes using Wiener and neural network models is compared for several different processes and for three different types of input signal: a pseudorandom binary signal (prbs), an inverse-repeat pseudo-random binary signal (irprbs) and a multisine (sum of harmonics) signal. Experimental results on an electronic nose are presented to illustrate the applicability of the techniques discussed.