Viscosity based liquid classification using Artificial Neural Network

Every liquid has certain properties, which signifies its usage in various applications. One of the major property for identification of liquid is through the determination of its viscous behaviour. This paper aims at classification of different liquids by comparing their viscosity. A hardware has been designed to obtain the frequency of different liquids. A strategic relationship between the frequency of liquid and their viscosity has been significantly correlated by using Artificial Neural Network (ANN) model. A model has been designed using ANN which is a viable tool for classification of 49 different liquids on the basis of their frequencies. This helps in categorizing these liquids into 2 categories i.e. type 0 and type1. Promising results were obtained by the ANN model for different categories of liquids. High correlation coefficient 0.96521 and less Mean Square Error (MSE) 0.0145 were achieved which provides a framework for liquid identification based on their viscosity.