With the improvement of discharge standards of sewage treatment plants and the proposal of"double carbon",related technologies of smart water have developed rapidly in recent years,among which the research and practice of aeration control system is the most important part.Based on the main goals of effluent quality,energy consumption and the target point of am-monia nitrogen,this paper developed a multi-objective aeration control system combining activated sludge model and machine learning to break the conventional aeration control mode of"constant DO"based PID control.The system was deployed and operated in Yinchuan No.1 recycled water plant,and on the basis of ensuring stable effluent quality and reaching standard,the workload of the central control room was reduced by 2/3,the average aeration power unit consumption was re-duced by 12％,and the dosage of phosphorus removal agent decreased by 30％.
维普
万方数据