Application Of Data-driven Control To A~2O Plant Operation

Activated sludge process (A~2O) is a widely use way in the sewage treatment. Thismethod is famous for its simple process, widely use and won’t cause the secondarypollution. However，with its strong coupling and large time-delay, activated sludgeprocess is hard to be modeled and controlled. Based on the subject “Switched DynamicSystems Modeling and Multiple-Objective Optimization Control of Activated SludgeProcess” which comes from the Urban water resources and water environment of statekey laboratory, we designed the black box controller for the activated sludge model1(ASM1) to improve water purification ability. Rid from the object model and prior data,this method only base on on-line data. Simultaneously it can help to restrain disturbanceand save energy and resources. First, this article analyses the principle of the ASM1.Based on its mathematical model, the simulation of the ASM1is done. In the aftersimulation, it provides data for the system. Secondly, we present the algorithm of datadriven control, especially Iterative Feedback Tuning. Also we analyze the feasibility ofits industrial applicability in waste water control. Thirdly, we write the IFT’s programvia Matlab. Finally, a design of ASM1lab monitoring and control system has beenpresented. The key feature of the paper is to try a new method in waste water control. Itcan help to restrain disturbance and save energy and resources. All these work will havea promising application in the practical waste-water treatment.