Artificial Intelligence Based Alum Dosage Control in Water Treatment Plant

Supplying good quality of drinking water is a challenging task during the rainy season and floods. During this period water becomes highly polluted with suspended solids which increase the water turbidity. Alum is used to reduce the turbidity of the water. Typically in water treatment plants alum dosage is decided by the Jar test and the desired alum dosage is added manually. This research proposes an automatic alum dosage mixing process. The alum dosage is controlled by an intelligent controller which consists of a dosage predictor, an inverse model of the dosage pump and a Pulse Width Modulation (PWM) controller. The optimal alum dosage is predicted by the dosage predictor. The PWM controller controls the flow rate of the alum dosing pump. This proposed method has been implemented in a laboratory based water treatment plant and it ensures the automation in water treatment plant to supply good quality drinking water. Keyword-Chemical dosing, Dosage predictor, Feedforward controller, Pulse width modulation, Water treatment plant