基于神经网络优化的低负荷生物质气掺烧锅炉数值模拟研究OA

[Objective]To investigate the effects of biomass gas co-firing on combustion stability and in-furnace parameters under low-load conditions,a 660 MW tangentially fired boiler was taken as the research object to carry out the study.A stability index was proposed,and a combined approach of numerical simulation and artificial neural network(ANN)was employed.[Methods]Comparative analysis was conducted between pure coal and co-firing conditions at loads of 100％,70％,50％,and 30％.[Results]The results show that the deviation of the temperature stability coefficient(MT)under co-firing is less than 3.9％,indicating stable combustion at low load conditions.When the unit load decreases from 100％to 70％,the average temperature in the main combustion zone drops by 147.87 K for pure coal and 69.37 K for co-firing,indicating a slower temperature decay.At 30％load,the NOx volume fractions in the reduction and burnout zones under co-firing are 0.051 2％and 0.044 4％,which are lower than 0.093 3％and 0.078 6％under pure coal combustion condition,with smaller fluctuations of other parameters.Furthermore,an artificial neural network(ANN)model was developed to describe the complicated relationships among in-furnace parameters,and the results show that the regression coefficients R2 for temperature,CO2 volume fraction,and NOx volume fraction predictions are all greater than 0.96 in both pure coal and co-firing conditions.[Conclusion]This study provides support for optimization and prediction of low-load operation in biomass gasification co-firing boilers.