A self learning algorithm based on data value lifecycle model for the accurate chemical dosing of wastewater treatment

Abstract

Abstract In the process of wastewater treatment (WWT), the instability and hysteresis of the chemical dosing cannot make sure the stability of the water quality. At the same time, chemicals were wasted or overdosed in the process. In this study, we proposed a self learning algorithm, which based on the regression method modified using the data value lifecycle model to computer the accurate quantity of chemical dosing. The mechanisms of discard and cultivation data were established to make it alive in the algorithm. The algorithm is self learning according to the wastewater characteristics as the time goes on using the mechanism. We can make sure the quality of the wastewater is stability and economical by using the artificial intelligence. The experiment approves that the artificial intelligence algorithm was useful and economical to chemical dosing of wastewater treatment.