Research On Key Technologies For Optimal Control Of Urban Sewage Treatment Process Based On Hierarchical Control

The growing size of cities has led to the continuous improvement of sewage treatment capacity and stricter sewage discharge standards.At present,there are many problems in urban sewage treatment plants,such as high energy consumption,low operation efficiency,complex inflow flow,substandard effluent quality and so on.Therefore,in order to obtain better control performance of urban sewage treatment process,this paper studies the key technologies of optimal control of sewage treatment process.The details are as follows:1.Considering that in the process of urban sewage treatment,the mechanism model is complex and difficult to establish,and only the input and output data are known,this paper mainly models the sewage treatment process based on subspace identification method.Specifically,it includes the state space description of the system,N4SID subspace identification,and the determination of operating variables and controlled variables.Finally,the state space model of sewage treatment process is obtained,which lays the foundation for the subsequent optimization and control work.2.It is considered that the urban sewage treatment process contains coupled parameters and is a multi input multi output nonlinear system.Combined with the state space model identified in Chapter 2,a double-layer structure model predictive control(MPC)with multi priority steady-state target calculation is designed,and the steady-state targets of dissolved oxygen(DO)and nitrate nitrogen(NO)are calculated and obtained;Further,combined with dynamic control,the target values of do and no are controlled within the expected range.The simulation results show that the accuracy of double-layer structure model predictive control is higher than that of PI control.3.Considering that it is difficult to control DO and NO in the process of sewage treatment,the traditional control strategy can not achieve good control accuracy for nonlinear and timevarying objects.Meanwhile,long short term memory(LSTM)network has unique advantages in dealing with nonlinear and time-varying problems.In this paper,the modeling and control of sewage treatment based on LSTM network are designed,and combined with the steadystate target calculation results in Chapter 3,the optimal control of sewage treatment process is realized.The simulation results show that compared with PI control and MPC,the controller based on LSTM network can reduce the maximum deviation of DO and NO concentration by 72.1%,28.1%and 85.3%,67.5%respectively,and the optimal control effect is obvious.In this paper,the urban sewage treatment process as the research object,based on subspace identification,multi-priority steady-state goal calculation,optimization control,LSTM network control and other technologies,to carry out the research of multi-objective optimization control method,in order to achieve the sewage treatment plant energy conservation and emission reduction,stable operation,effluent quality standards.