Optimization control method for combustion of quadrangular tangential boiler based on optimization of air distribution mode

Abstract

The invention discloses a four-corner tangential circle boiler combustion optimization control method based on the optimization of the air distribution mode. First, offline modeling: screening historical data and performing data fitting, and establishing an initial model of a support vector regression (SVR); and then performing online calculation: using The accurate online support vector regression method (AOSVR) updates the model, solves the nonlinear programming according to the updated model and the set constraints, and obtains the optimal control quantity. Aiming at the problem that there are many control variables in the four-corner tangential boiler, the present invention is based on the idea of ​​overall adjustment of the air distribution method, and uses the shape parameter instead of the control quantity through the simplified method of data fitting, and optimizes the distribution method of the control quantity as a whole, thereby avoiding the occurrence of machine learning The problem of "curse of dimensionality" is solved, and the calculation speed is improved while ensuring the accuracy. The optimized control strategy combined with this simplified method can improve boiler efficiency or reduce nitrogen oxide emissions on the premise of ensuring safe and stable operation of the boiler.

Description

technical field [0001] The invention relates to a combustion optimization control method for a four-corner tangential circle boiler based on the optimization of the air distribution mode, and belongs to the technical field of thermal automatic control. Background technique [0002] In recent years, with the gradual improvement of environmental protection requirements, power plants need to reduce pollutant emissions as much as possible while ensuring combustion efficiency and safety. Therefore, the optimization of boiler combustion process has increasingly become a research hotspot. At present, the combustion optimization mainly adopts the data-based modeling method, that is, the nonlinear model of the boiler combustion is established first, and then the performance indicators related to the boiler efficiency and nitrogen oxide (NOx) emission are constructed, and the operating parameters are optimized using the optimization algorithm. Optimize calculations. Therefore, estab...

AI Technical Summary

Problems solved by technology

In the existing work, all the state parameters and controllable variables are used as the input variables of the model when modeling the combustion system. 20-30 input variables, falling into the "curse of dimensionality problem"

As the dimension of the data set increases, the number of samples required increases exponentially, and the sparsity of the data will also increase, resulting in slower calculation speed and lower accuracy of the algorithm, which will lead to online update of the later model and online optimization of the combustion optimization problem. Solving caused great difficulty

[0004] On the other hand, the existing work generally does not update the model online, and the boiler combustion system is greatly affected by interference factors such as changes in coal quality, and must be updated when the model mismatch is serious to ensure the accuracy of the model

Images