Modeling of thermodynamic properties using neural networks - Application to refrigerants

Thermodynamic data on environmentally-acceptable refrigerant fluids are of the utmost interest for the refrigeration industry and, in particular, to design and optimize refrigeration equipment such as heat exchangers and compressors, Up to now, the simultaneous representation of vapor-liquid-equilibrium (VLE) and pressure-volume-temperature (PM data is not satisfactory enough with respect to experimental accuracies. New models are then highly required. Therefore, an effort has been made in our laboratory to develop an alternative to a simple equation of state. This work deals with the potential application of artificial neural networks (ANN) to represent PVT data within their experimental uncertainty, and then obtain, by numerical derivatives, other thermodynamic properties such as enthalpies. entropies and heat capacities. Tests presented were performed an data corresponding to some refrigerants from 240 to 340 K at pressures up to 20 MPa. (C) 2002 Elsevier Science B.V. All rights reserved.