ZHONG Yonglu, LI Haishan, LIU Fasheng, et al. PCA-SVR model based NOx emissions prediction for coal-fired boilers[J]. Thermal power generation, 2015, 44(1): 87-90.
ZHONG Yonglu, LI Haishan, LIU Fasheng, et al. PCA-SVR model based NOx emissions prediction for coal-fired boilers[J]. Thermal power generation, 2015, 44(1): 87-90.DOI:
Taking a 700 MW ultra-supercritical boiler as the object
a nitric oxides(NOx)emissions model was established
by combining principal component analysis(PCA)and support vector regression(SVR).The PCA was employed to analyze 2 000 operational records of the boiler
which reduced the dimension of the input vectors and kept the most information of the data simultaneously.The SVR was used to build the NOxemissions prediction model by taking the principal components as inputs and NOxemissions as output.Compared with the conventional SVR model
the PCA-SVR model required less computational cost and had a higher predictive accuracy.According to the test
the average relative error between the measured value and the result predicted by the PCA-SVR model was below 1%.Furthermore