Predictive Control Of Nitrogen Oxide Emissions From The Denitration System Of Coal-fired Boilers

As the country attaches more and more importance to environmental protection,the state has actively promoted the ultra-low emission retrofit of thermal power plants,which are the main source of atmospheric pollution,and has accordingly formulated increasingly stringent air pollutant emission standards.SCR denitration technology is currently the main means of controlling NO_x emissions in thermal power plants.In the SCR denitration system,t he accuracy of data and the effectiveness of control strategies are the keys to achieving NO_x emissions standards for thermal power plants.However,due to the large inertia of the boiler combustion process and the on-site environmental impact,the detection data is delayed and inaccurate;due to the changing operating environment of the denitration system and many influencing factors,the traditional PID control quality is not ideal.How to solve the problems of hysteresis and non-linearity of the denitration system,reduce the fluctuation of NO_x concentration at the outlet,and reduce the cost of ammonia injection.This article has done the following three aspects:First,a soft-sensing and prediction model for the NO_x concentration at the inlet of an SCR denitration system based on a neural network is established.After analyzing the influencing factors of the NO_x concentration at the inlet of the SCR denitration system,the historical data of 330 MW units in a power plant were collected;the input variables of the model were determined after the correlation analysis,and the BP neural network model was established after necessary processing of the data;The genetic algorithm was used to optimize the BP neural network.The output of the model was compared wi th the measured data to verify the validity of the model.Secondly,in-depth study of dynamic matrix control algorithm,a dynamic matrix control algorithm(FF-DMC)based on feed-forward signal is proposed.After a simulation comparison of a model,it is verified that the dynamic matrix control has an efficient set value tracking capability,and the feedforward dynamic matrix control has a good anti-interference performance.Finally,it is proposed that it can be applied to large delay and large inertia syst ems.Thirdly,the parameters of the model were optimized by the method of particle swarm parameter optimization,and the controlled object model of the SCR denitration system was established.The model output was compared with the measured data to verify the validity of the model.Based on the identified model,the design A dynamic matrix controller(DMC)was obtained,and a good control effect was obtained through simulation.The inlet NO_x concentration was introduced as a feedforward,and a feedforward dynamic matrix controller(FF-DMC)was designed.The simulation showed that the FF-DMC control system can Enhance the anti-interference performance of the control system.