Application of Artificial Neural Network for Prediction of Liquid Viscosity

Abstract

Viscosity is one of the important quality properties of fluids. It is a measure of the resistance of a fluid to being deformed by either shear or extensional stresses. Change in viscosity with temperature is very important in engineering calculations. In the present work a new method based on artificial neural network (ANN) for prediction of the viscosity of a compound has been proposed. The required data were taken from Perry's Chemical Engineers' Handbook. The accuracy and trend stability of the trained networks were tested against unseen data. Different training schemes for the back-propagation learning algorithm, such as scaled conjugate gradient (SCG), Levenberg-Marquardt (LM), and resilient back propagation (RP) methods were used. The SCG algorithm with 19 neurons in the hidden layer was found to be the most suitable algorithm with minimum mean square error (MSE) of 0.000381. Results of the ANN model were evaluated against the unseen data and then compared with the empirical models. The ANN's capability to predict viscosity is one of the best estimation methods with high performance.