Research On Modeling And Operation Optimization For Combustion Process Of Thermo-Electric Boiler

The process of boiler combustion is to achieve the realization of automatic control,ensureing its implementation in a stable and safe state,which actulizes energy conversion of chemical energy into heat energy.Meanwhile,the process also should be optimized,so as to can make the entire unit maintain a good benefit in the process of operation.With the further requirement of the control of pollutant emission in thermal power plant,the combustion process should not only pay attention to the economic benefit,but also achieve the environmental protection index.Therefore,the combustion process of boiler is multi-objective control and operation optimization of mutual correlation and coupling.In order to solve the problem of the optimization and control of the process better,this paper studies from the following three aspects.(1)The method of data integration.In this paper,the integration technology based on data driven modeling is studied.The collaborative filtering(CF)recommendation engine algorithm is introduced to complete the sample,which the low calorific value is missing in the boiler combustion.Meanwhile,using principal component analysis to deal with the high dimensional problems of samples.Clustering algorithm and statistical method are used to analyze the data of the two dimensional samples.Combined with the advantages of both methods,FCM-MD algorithm is proposed to identify the outliers in the combustion data set,which can effectively distinguish the center outliers and edge outliers.(2)Modeling study on thermal efficiency and NOx emission of thermal boiler.According to the characteristics of boiler combustion,which is a nonlinear and multi-conditional and multi-coupling process,a multi-model modeling method based on the combination of fuzzy C mean clustering,least square support vector and weighted connection(FCM-LSSVM-WC)is proposed.This algorithm was used to research the modeling of boiler thermal efficiency and NOx emissions,which can classify boiler combustion conditions in the process of modeling,and according to the same rules of multi-model output connection,ensuring the modeling process of the consistency and uniformity of the modeling process.The all above methods have established is to realize the optimization of the thermo-electric boiler combustion.(3)Study on optimization and control of combustion process of thermo-electric boiler.With reference to the idea of optimization and control integration,the output variables of combustion process control can be maintained as much as possible at the given level of optimization.Using improved quantum genetic algorithm(QGA)to iterative optimize the multi-model of boiler thermal efficiency and NOx.After the optimal set value of flue gas oxygen content in a certain range of load is obtained,model predictive control is used to track the value in combustion control.In the meantime,using interval control of main steam pressure and furnace pressure to make the process of boiler combustion in a stable and safe state,which makes the optimization and control to achieve the optimal target.