Modeling And Optimization Guidance Of Boiler Combustion Characteristics Based On Data Driven

As an important large-scale heat production equipment in the heating industry,the boiler is difficult to achieve real-time accurate adjustment due to the complex working conditions and large lag,which makes it difficult to maintain in an efficient state.Therefore,it has important research value to design practical and effective scheme to complete the boiler combustion optimization guidance.With the development of science and technology,a large number of field data are acquired and stored.The data-driven method provides a new idea and possibility for combustion optimization guidance.Therefore,the guide topic of modeling and optimization of boiler combustion characteristics based on data is established.The main contents are as follows:First of all,the selection of relevant input variables is the premise of building the boiler combustion characteristics model.In view of the problem that it is difficult to select the variables that affect the boiler efficiency due to the delay disturbance,the time delay calculation method based on cross correlation and Pearson correlation analysis method are combined to obtain the variable correlation coefficient considering the delay.The influence of the input variables without considering the lag and considering the lag on the prediction effect of the model is compared.The experimental results show that the latter has better prediction effect.Therefore,the input variables of boiler combustion characteristic model are determined based on the results of correlation analysis and operation experience.Secondly,the establishment of accurate boiler combustion characteristic model is the basis of boiler combustion optimization guidance.Aiming at the problem that the combustion process of boiler is complex and nonlinear,and it is difficult to model by mechanism,the least square support vector machine(LSSVM)method of particle swarm optimization(PSO)is used to model,and it is found that the prediction accuracy of the model is low.Therefore,the improved particle swarm optimization least square support vector machine method is used to model,which can effectively improve the prediction effect of the model.This method can establish a more accurate static combustion characteristics model.In order to solve the problem that the model can not accurately describe the future characteristics with the change of production process,the sliding time window method and the European distance method are used to carry out the model updating comparative experiment.It is found that the former can only describe the latest characteristics on the time axis,while ignoring the influence of the effective characteristics in the history on the model;the latter can not only retain the effective characteristics in the original model,but also add new system characteristics according to the set threshold value,and reduce the calculation amount of model updating.Through the nonlinear function experiment and the boiler combustion characteristic modeling experiment,it shows that the model based on Euclidean distance updating method can describe the combustion characteristic to the maximum extent,and it has better effectiveness and applicability for the boiler combustion system with variable working conditions.Finally,the purpose of establishing accurate boiler combustion characteristic model is to serve for optimization guidance.Based on the updated model of boiler combustion characteristics by the Euclidean distance method,the improved particle swarm optimization algorithm is adopted to optimize the combustion efficiency by adjusting the given value of adjustable variables in the current state.The simulation results show that the boiler efficiency can be increased by 1.84%on average,which proves that the method has certain effectiveness and guidance.Therefore,a boiler combustion optimization experimental platform based on open database connectivity(ODBC)is designed,which can provide real-time auxiliary operation suggestions for the operation and decision-making of the heating production site.