Research On Combustion Optimization Of Utility Boilers Based On Adaptive Fuzzy Method

In the background of energy conservation,the utility boiler combustion optimization technoloy is one of the effective method and approach for energy saving and emission reduction in the power plant,which is paid close attention from power plants.According to the special operation feature of the domestic coal-fired power unit,the paper researches mainly the key issues of the combustion optimization technoloy,such as the modeling and optimation of combustion system.The main work and research results are as follows:1)The existing methods are elaborated;considering the shortcomings of the existing methods,a new efficient modeling method,named adaptive fuzzy method based on the tree structure which is abbreviated as fuzzy tree(FT),is studied carefully.Then,a robust fuzzy tree,named ?-fuzzy tree(?-FT),is introduced.2)Considering the strong correlations and coupling of input variables in thermal power process,partial least squares(PLS)method is applied to extract important variables information and select variables for input thermal data of model.Subsequently,the obtained feature matrix is used as the input of ?-FT to establish the PLS-?-FT model of combustion system.The model is compared with other modeling methods.The results show that the variable correlations is eliminated,the model complexity and the number of the variable dimensions are decreased,and the prediction accuracy and generalization ability of the model are enhanced through PLS variable selection.3)Considering the noise and outliers of input variables in thermal power process,the paper proposes the weighted fuzzy tree(W-FT)based on the local outlier factor(LOF),and two typical nonlinear examples are used to validate the proposed W-FT.Then,the combustion models based on W-FT are established,which are made a comparison with other modeling methods.The results show that the proposed W-FT can effectively recognize the noise and outliers;the built models have higher prediction accuracy and stronger generalization capacity.4)In order to improve energy efficiency of boiler unit and reduce pollutant emissions,the exhaust temperature model and NOx emissions model are built based on ?-FT,and two types of optimization strategies are proposed based on the exhaust temperature model and NOx emissions model to optimize the adjustable parameters within a certain range by using the modified fruit fly optimization algorithm(MFOA).Simulation results show the proposed models have higher prediction accuracy and stronger generalization capacity;the proposed two optimization strategies can achieve to reduce exhaust temperature model and NO_x emissions;the proposed combustion optimization program spends less time,and are suitable for online applications,which serve as a important reference for the actual operation in utility boiler.