Research On Thermal Power Optimizing Operation Based On Data Mining And Prediction Of Proximate Analysis

Power plant boiler operation optimization is an important method to keep the unit operating safely and economically.The control system of boiler which includes physical and chemical processes is so complex that make the boiler operation optimization existing many difficulties.With the development of information technology,the SIS system of power plant has accumulated a lot of data.Finding the best way from the boiler operating data to improve thermal efficiency has practical significance and research value.However,due to the diversity of the data,artificial analysis is almost impossible.In this paper,the data mining theory is combined with boiler operation optimization,conducting a preliminary research about the optimization of power plant boiler based on data mining.This paper focuses on the application of data mining technology in the power plant boiler operation optimization.The main contents of this paper are as follows.It is necessary to improve the quality of boiler operation data for providing the basis for modeling.Attributes of boiler operation data which exists abnormal and null values is so large that make the quality of data poor.Therefore,it is need to deal with data before using the operation data to establish model.The processing mainly includes cleaning data,the extraction of steady conditions and the selection of data.Data preprocessing is a key step in the data mining technology which can effectively provide guidance for boiler operation optimization.In power plant,the coal information is obtained offline.It is difficult to guide the operation of the boiler effectively.Based on the element online detection technology,an model of predicting proximate analysis based on elemental analysis is proposed,which provides a solution for coal quality on-line inspection in the future.The model of predicting NOx by using the coal quality data and the boiler operating data is also established,which realizes real-time control of NOx emission.Based on the results of data processing and the model of predicting NOx emission,the parameters of NOx and fly ash are used as optimization parameters.Fuzzy association rule mining algorithm is used to obtain the relationship between NOx emission,unburned carbon content in the fly ash and flue gas oxygen content to obtain the best optimal value of oxygen content which makes the NOx emission and unburned carbon content in the fly ash at a low level.