Boiler wide-load NOx emission concentration prediction method based on model migration

Abstract

The invention discloses a boiler wide-load NOx emission concentration prediction method based on model migration. A numerical simulation calculation center, a base model training center, an operationdata communication interface, a base model migration and updating mechanism and a NOx emission prediction model communication interface are included. The method comprises: in a feasible region of mainoperation parameters of a boiler, obtaining a boiler full-working-condition sample under a designed coal type through off-line simulation by taking the designed coal type as a reference, and establishing a NOx emission prediction base model by adopting machine learning; considering non-design coal types for boiler combustion, obtaining a small number of typical working condition samples through off-line simulation, then migrating the design coal type base model to the non-design coal type working conditions by adopting Gaussian process regression, and forming a base model library adapting tothe change of multiple coal types; considering the difference between the actual working condition and the simulated working condition, selecting a base model according to the actual coal quality, andthen projecting the base model to the actual operation state of the boiler through transfer learning based on the operation data, so as to realize accurate prediction of the NOx emission concentration under the wide load of the boiler.

Description

Technical field [0001] The invention belongs to the field of thermal power generation, and in particular relates to a method for predicting the NOx emission concentration of a boiler at a wide load based on model migration. Background technique [0002] NOx is one of the main pollutants generated by the combustion of coal-fired boilers. The accurate prediction of the NOx emission concentration of the boiler can provide decision support for the adjustment of the boiler operation mode, so as to meet the increasingly stringent environmental protection requirements. However, the combustion of pulverized coal in the boiler is a complex process of multi-field coupling, and there are many factors affecting the concentration of NOx emissions. In addition, as the penetration rate of new energy continues to increase, coal-fired units are more involved in deep peak shaving tasks, and boilers are often operated at ultra-low loads that deviate from design conditions, which brings new ideas to...

AI Technical Summary

Problems solved by technology

However, since the units with deep peak regulation often operate under ultra-low loads that deviate from the design conditions, the data sample coverage is small, and the applicability of the model is difficult to guarantee

In order to solve the above problems, the patent CN107726358A adds CFD simulation samples to the modeling samples to expand the coverage of the samples, but there is a certain difference between the CFD numerical test results and the actual operating characteristics of the boiler, so the obtained model is difficult to accurately predict the NOx emission concentration

Although the NOx emission concentration prediction method disclosed in the patent CN109670625A can realize online updating of training samples to adapt to changes in boiler operating characteristics, it does not consider the influence of coal quality characteristics on NOx emission concentration, so the applicability of the method is limited

Images