Dynamic Score Normalization Technique using Mahalanobis Distance to Predict the Level of COD for an Anaerobic Wastewater Treatment System

Worldwide, environmental regulatory establishments are setting strict standards for discharge of wastewaters from industries. Anaerobic digestion is an environment friendly technology to treat industrial wastewaters. Anaerobic digestion is a series of processes in which microorganisms break down biodegradable material in the absence of oxygen. For the treatment, cheese-dairy wastewater is taken. Chemical Oxygen Demand (COD) is an essential parameter for determining the pollution level of effluents prior to discharge. The whole procedure in an anaerobic wastewater treatment system for the prediction of COD includes data collection and pre-processing (statistical normalization techniques), feature selection and prediction of COD using Back Propagation Neural Network (BPN). At the time of the pre-processing several normalization techniques are used. The Dynamic Score Normalization technique with Mahalanobis Distance (DSN-M) is proposed which is compared with other existing normalization techniques. BPN is found to be the best for the prediction of COD. In order to determine the effect of proposed (DSN-M) technique in predicting the COD level, experiments were carried out to find the prediction of accuracy of BPN.