Artificial neural network prediction of exhaust emissions and flame temperature in LPG (liquefied petroleum gas) fueled low swirl burner
Loading...
Date
2013-09
Journal Title
Journal ISSN
Volume Title
Publisher
Energy by Elsevier
Abstract
This study deals with ANN (artificial neural network) modeling of a swirl burner. The model was used to
predict the flame temperature and pollutant emissions (CO (carbon monoxide) and NOx (nitrogen oxide))
from combustion of LPG (liquefied petroleum gas) in the swirl burner. The data for the training and
testing of the proposed ANN was obtained by combusting LPG at various equivalent ratios (LPG/air ratios)
and swirler’s vane angles in a low swirl burner. Vane angles of 35e60 in steps of 5 and equivalent ratios
of 0.94, 0.90, 0.85, 0.80, 0.75, 0.71, 0.66 and 0.61 were considered. An ANN model based on standard
back-propagation algorithms for the swirl burner was developed using some of the experimental data for
training and validation. The performance of the ANN was tested by comparing the predicted outputs
with the experimental values that were not used in training the network. R values of 0.94 were obtained
for CO and NOx and 0.99 for flame temperature. These results show that very strong correlation exists
between the ANN predicted values and the experimental results. Therefore, this study demonstrates that
the performance and emissions of swirl burner can be accurately predicted using ANN approach.
Description
Staff Publication
Keywords
ANN, Swirl burner, Liquefied petroleum gas, Emissions, Flame temperature
Citation
Adewole, B. Z., Abidakun, O. A., & Asere, A. A. (2013). Artificial neural network prediction of exhaust emissions and flame temperature in LPG (liquefied petroleum gas) fueled low swirl burner. Energy, 61, 606–611. doi:10.1016/j.energy.2013.08.027