The BP Network Model for the Control of the Coagulation and the Filter Backwashing System

The optimization of coagulant dosage in drinking water treatment has become an issue of particular concern for utility managers. We studied the shortage of the operational control methods. And this paper presents the application of ANN to model the prediction of coagulant dosage. The model is constructed based on the database of operating parameters and water quality. The structure and the parameters of the neural networks for engineering control are obtained through the optimization algorithm. The present model is building for the coagulant dosing system and filter system prediction and control. And this model realized in the Matlab environment. It is applied to the engineering operation and the field experiment. Results show that the BP networks model can predict the coagulant dosing rapidly and effectively. Compare with the ordinary methods, use this BP model can reduce the operational cost.