Regression model with artificial neural network for anaerobic digestion of wastewater treatment

Regression analysis can be used to model the relationship between predictor and response variables and is a good choice when all the predictor variables are numeric and continuous valued. In this paper, multilayer perceptron neural network is used for predicting the experimental values obtained in a laboratory scale system of anaerobic tapered fluidized bed reactor (ATFBR). The system study is the anaerobic digestion of synthetic wastewater derived from the starch processing industries. The input parameters considered for modeling are flow rate, CODin, pHin and hydraulic retention time. The output parameters are biogas yield and pHout. The Mean Square Error (MSE) obtained for the test dataset obtained with experimental set-up is as low as 0.1416.