Research On Intelligent Control Strategy Of SNCR Denitration System In 300MW Circulating Fluidized Bed Unit

Taking the plant's joint SNC-SCR denitration system of a certain power plant 300 MW circulating fluidized bed unit as the research object,through the analysis of the mechanism of NOX formation of furnace and influence factors,SNCR and SCR denitration mechanism and influence factors,and according to the current status of experimental unit denitration system control,the key to realize the unit automatic removal control of NOX is to solve the issues of SNCR denitration automatic control.NOX removal process of the SNCR denitration system with large delay,large inertia,multi-factor coupling features,at the same time,CFB boiler using SNCR denitration system has the characteristics that the production of NOX can not be measured,for this kind of the controlled object,conventional control method often can't meet the control requirements,and some of the advanced control that can't establish a more accurate mathematical model of the object is difficult to be used in the actual implementation,in this case,according to the theoretical knowledge combined with the experience of production personnel knowledge,combining multiple control theory with expert knownedge in related fields,put forward a kind of intelligent control strategy for SNCR denitration that is suitable for implementation,the main contents include:1.Choosing the BP neural network with the characteristics of strong nonlinear approximation ability,easy to implement and so on,According to the relationship between the produced mass concentration of NOX and the load and the wind coal ratio without using the SNCR system,as well as the relationship between the measured mass concentration of NOX and the load,the ratio of air to coal and the amount of ammonia with using the SNCR system,establishing the two kinds of BP neural network model,and studying the concrete method of realizing the BP neural network in xinhua DCS,then the trained network model was writed in DCS configuration program,it realized the online prediction in mass concentration of NOX production and mass concentration of the left NOX after reduction reaction.To solve the NOX automatic control problem of SNCR system provides a more accurate model,and provides the conditions for the optimization of the denitration automatic control and subsequent implementation of the control strategy.2.A SNCR control strategy based on intelligent control is designed,the basic structure of control strategy is based on expert control,and then combined with fuzzy control,BP neural network and model free adaptive control algorithm,at last,the eight original structures are formed,including: Setting point formation structure based on expert experience,Hierarchical optimization structure,the interval control based on expert control,Steady state optimization based on model-free adaptive,the feedforward control based on BP neural network,the reductant low/high limit protection based on the BP neural network,the reductant low limit adaptive protection based on expert control and NOX fast protection based on the fuzzy control.The intelligent control strategy is combined with the Xinhua DCS system,after debugging and running,Finally,the control problems caused by the large delay,nonlinear,large inertia and multi variable coupling in the NOX removal process are solved,realizing the automatic operation of 100% rate of denitration system in the test unit,and ensure the safe,stable and economical operation of denitration system.This topic research results have been successfully applied to the two 300 MW CFB unit of denitration system,application effect is remarkable and has a certain application prospect.