Modeling of Chlorine Decay using Artificial Neural Network

Abstract

In this paper an attempt is made to model chlorine decay using Artificial Neural Network (ANN). Initial chlorine concentration, fast and slow reacting organic and nitrogenous compounds and reaction rate constants of the compounds are used as inputs to the ANN model and the chlorine decay at different points in the decay curve are evaluated. ANN is trained by two different methods namely single output model and multi output models. Predicted data are compared with observed using correlation coefficient. Result indicates multi output model able to model more accurately than single output model.