Modelling of nonlinear state-space systems using a deep neural network

Julian Parker; Fabian Esqueda; André Bergner
DAFx-2019 - Birmingham
In this paper we present a new method for the pseudo black-box modelling of general continuous-time state-space systems using a discrete-time state-space system with an embedded deep neural network. Examples are given of how this method can be applied to a number of common nonlinear electronic circuits used in music technology, namely two kinds of diode-based guitar distortion circuits and the lowpass filter of the Korg MS-20 synthesizer.