Optimizing combustion efficiency of a circulating fluidized boiler: A data mining approach

Abstract

A data mining approach was applied to analyze relationships among 54 parameters of a circulating fluidized-bed boiler. Knowledge was extracted from the data by machine learning algorithms. The extracted knowledge was used to determine ranges of process parameters (control signatures) that led to the increased efficiency of the combustion process. The research has shown that the efficiency can be predicted to the same degree of accuracy with and without the data describing the fuel composition or boiler demand levels. This discovery might have profound impact on the research directions in optimization of the energy production. Adjusting parameters of the control system has led to improved efficiency of the combustion process. The proposed data mining approach is applicable to different types of burners and fuel types. It is well suited to perform tradeoff analysis between various performable measures, e.g., efficiency and emissions.