Neural network-based prediction of the oscillating behaviour of a closed loop thermosyphon

The aim of this work is to address the problem of modelling the dynamical behaviour, manifested during unstable operation, of an experimental closed loop thermosyphon. A generalised Nonlinear Auto-Regressive Moving Average with eXogenous inputs (NARMAX) model, implemented by means of neural networks, is used to address the identification of the system dynamics by means of input–output experimental measurements.