Research On Data Modeling And Control Method Of Flue Gas Denitrification System In Coal Fired Power Plant

With the rapid development of economy and society,environmental pollution has increasingly become a prominent problem in China,especially the air pollution problem led by haze and acid rain is the most serious.As the main source of atmospheric pollution,NOx emission control by SCR flue gas denitration technology is widely used in coal-fired power plants.Although the traditional SCR flue gas denitration technology can play a certain role,the accurate measurement of the NOx concentration at the inlet and the fast and stable NOx concentration at the outlet cannot be set at the set value,which cannot meet the actual needs of the site and cause waste of ammonia,and produce secondary pollution.With the introduction of NOx emission standards,it has important research value to study SCR denitration technology,ensure denitration efficiency and improve the economic benefits of the power plant.This article takes the normal operation of variable-load operation of a coal-fired power plant as the background,and takes the SCR denitration system as the research object to study the accurate measurement of the NOx concentration at the system inlet and the rapid stabilization of the NOx concentration at the set value.Soft measurement is used instead of traditional hardware measurement,data modeling is used to accurately predict the NOx concentration at the SCR inlet,and advanced control algorithm is used to optimize the system's ammonia injection.The main research contents are as follows:1.This paper introduces SCR denitration technology and application and other related theoretical knowledge,the status quo and existing problems of flue gas denitration in this power plant.Then selects the influencing factors of the system in the massive database of DCS,and uses k-nearest neighbor algorithm to select the factors that have the strongest correlation with the concentration of NOx at the outlet as the input variables of the model.2.Aiming at the strong dynamic and weak nonlinearity of SCR denitration system,a soft measurement model based on time difference partial least squares is established to realize the accurate measurement of NOx concentration at the entrance,the model has strong generalization ability and fast calculation speed,and is suitable for online use in engineering3.For higher-order dynamic systems and dynamic uncertain systems,dynamic models based on PLS external model expansion and internal model expansion are established respectively,use the fuzzy curve method to analyze the correlation between variables,determine the time delay of each variable in the input and output matrix,and use an adaptive method to continuously update the model.By calculating and comparing the model accuracy evaluation indicators of several modeling methods,the importance of the time delay of output variables for accurate modeling is illustrated.4.The traditional ammonia injection control strategy is introduced,and the PID control method is used to optimize the ammonia injection amount.After analyzing the problems of the current denitration control system of the power plant,advanced advanced predictive control method is proposed,the DMC predictive control method is used to optimize the ammonia injection amount of the system,and the NOx concentration at the SCR outlet was quickly stabilized at the set value.Compared with the PID control method,the advantages of the predictive control method are highlighted.