The prediction of chemical oxygen demand (COD) or suspended solid (SS) removal using statistical methods and the artifical neural network in the sugar industrial wastewaters

Özkan G., Akin B. a., ÖZKAN G.

ARPN Journal of Engineering and Applied Sciences, vol.8, no.12, pp.978-983, 2013 (Scopus) identifier


A static model and Artificial neural network for the prediction of COD or Suspended solid (SS) removal by chemical coagulation have been examined in a sugar industrial wastewaters and the performance of ANN model has been compared with the statistical model based on central composite experimental design. Three independent variables which effecting amount of COD or SS removal were selected. Namely are; quantities of the chemicals, pH and mixing rates. Amount of COD or SS removal were chosen as the dependent variable (target data). A second order statistical model has been considered to show the dependence of the amount of COD and SS removal on the operating parameters. A backpropagation ANN has been used to develop a model relating to amount of COD removal. It is observed that a neural network architecture having one input layer with three neurons, one hidden layer with three neurons, one output layer with one neuron and an epoch size of 48 gives better prediction. The predictions are more accurate than those obtained from regression models. © 2006-2013 Asian Research Publishing Network (ARPN).