Research On Optimization Of SCR Denitrification Based On Multi-model Predictive Control

As Chinese society pays more and more attention to the field of environmental protection,the monitoring of nitrogen oxide(NOx)emissions during the power generation process of thermal power plants has also received extensive attention.SCR(Selective Catalytic Reduction)is currently the most widely used method for controlling NOx emissions,and is one of the main research fields at present.However,the controlled object of the SCR flue gas denitrification system usually has the characteristics of large delay,large inertia and nonlinearity.Therefore,in practical engineering applications,the injection of ammonia injection is likely to be untimely during the denitration process,resulting in failure to control the outlet NOx Effective control has caused damage to the environment.Therefore,in this paper,the optimization research of SCR denitration control system is carried out,and the main research contents include:(1)A prediction model was established for the inlet NOx concentration of the SCR denitration system.In view of the large hysteresis of the traditional measurement of the inlet NOx concentration of the SCR system,and the simple single-step prediction cannot solve this problem well,this paper proposes a variational modal decompositionautoregressive differential moving average based on Model(VMD-ARIMA)for multi-step prediction models.First,select an appropriate decomposition level for the inlet NOx time series,and then perform VMD decomposition.Secondly,the order-determination criterion is used to determine the order of the model for each layer component,and then the ARIMA model is used for multi-step prediction,and the model is maintained online.Finally,the multi-step predicted values of each layer component after VMD decomposition are added to obtain the multi-step predicted value sequence of the inlet NOx.(2)A γ-incremental stepped generalized predictive control algorithm(γ-SGPC)is established.Aiming at using traditional control methods to control the SCR denitration system,which cannot achieve the required control effect,an SCR optimization control strategy based on γ-incremental stepped generalized predictive control is established.Firstly,the basic principle and structure of the γ-SGPC control method are introduced.Secondly,the influence of the parameters of the γ-SGPC controller on the control effect is analyzed,which provides the basis for the selection and identification of the subsequent parameters.Finally,the auxiliary loop of the PID control ammonia injection valve is added to build a cascade control system,and the anti-interference and tracking performance of the γ-SGPC controller are studied,and the performance of the γ-SGPC controller is better than that of the dynamic matrix controller(DMC).With the performance of the PID controller,it is more suitable for complex systems.(3)A multi-model predictive control strategy for ammonia injection feedforward is established.Considering that the relevant parameters of the controlled object of the SCR denitration control system will change with the change of operating conditions,and the use of a single traditional control method is not effective in controlling the problem.A multi-model γ-incremental stepped generalized predictive control strategy based on VMD-ARIMA predictive value feedforward is established.First,a γ-incremental stepped generalized predictive controller is established for the object models under three operating conditions.Then the fuzzy membership function is established,and the global controller is established through the fuzzy gain scheduling controller.Finally,the ammonia injection feedforward based on the VMD-ARIMA model is added to the control loop to form a feedforward-feedback optimal control loop.The simulation results show that the feedforward-feedback control system can accurately adjust the amount of ammonia injection,reduce the fluctuation of NOx at the outlet,and improve the control quality of variable working conditions.