M. Iatrou et al., Modeling of nonlinear nonstationary dynamic systems with a novel class of artificial neural networks, IEEE NEURAL, 10(2), 1999, pp. 327-339
This paper introduces a novel neural-network architecture that can be used
to model time-varying Volterra systems from input-output data. The Volterra
systems constitute a very broad class of stable nonlinear dynamic systems
that can be extended to cover nonstationary (tine-varying) cases. This nove
l architecture is composed of parallel subnets of three-layer perceptrons w
ith polynomial activation functions, with the output of each subnet modulat
ed by an appropriate time function that gives the summative output its time
-varying characteristics, The paper shows the equivalence between this netw
ork architecture and the class of time-varying Volterra systems, and demons
trates the range of applicability of this approach,vith computer-simulated
examples and real data. Although certain types of nonstationarities may not
be amenable to this approach, it is hoped that this methodology will provi
de the practical tools for modeling some broad classes of nonlinear, nonsta
tionary systems from input-output data, thus advancing the state of the art
in a problem area that is,widely viewed as a daunting challenge.