Design Of Denitration Control System And NO_x Concentration Prediction In Thermal Power Plant

In recent years,China’s economic strength is growing rapidly,the demand for electricity is also increasing,the number of newly built generating units is also increasing,and the generation of thermal generating units in the country’s total power generation occupies a dominant position.The flue gas emitted in the process of thermal power generation contains a lot of air pollution substances,among which nitrogen oxide(NO_x)is one of the main pollutants.Most of the ambient air quality indicators(PM2.5,PM10,SO2,NO2,O3 and CO)are related to NO_x.At present,the most widely used technology to control NO_x in power plants is selective catalytic reduction(SCR)denitration.With the strengthening of national control of NO_x emission from thermal power plants,it is essential to improve the denitration efficiency of thermal power units and reduce NO_x emission concentration.Based on the project of DCS control system for denitration of unit 1 and unit 2 of huaneng power plant in dezhou city,shandong province,with the aim of contributing to the denitration control of power plant,the denitration control system of dezhou power plant was studied and discussed by the method of establishing the prediction model of NO_x concentration.Firstly,the principle,process and steps of the selective catalytic reduction denitrification method were analyzed,and the important factors affecting the denitrification efficiency were summarized,which lays a foundation for the design,configuration and debugging of the interface and logic of the denitrification control system in huaneng power plant.Secondly,the network structure,software composition and hardware composition of hollysys MACS V6.5 system were analyzed,combined with the actual situation of huaneng power plant,the configuration of MACS V6.5system was studied,the main control equipment logic of huaneng power plant was summarized.Finally,for the problems of inaccurate measurement of NO_x concentration at the denitration outlet and large delay of automatic ammonia injection system during the trial operation of denitration control system in dezhou power plant,a single algorithm based on support vector machine regression(SVR)and a combined method based on empirical mode decomposition(EMD)and support vector machine regression have been adopted,to forecast the NO_x concentration of outlet of the denitration,through experimental simulation analysis,the combined EMD-SVR model has the highest prediction accuracy and can effectively predict NO_x concentration at the denitrification outlet.