Dual Wiebe Function Prediction of Eucalyptus Biodiesel/Diesel Fuel Blends Combustion in Diesel Engine Applying Artificial Neural Network

Paper #:
  • 2014-01-2555

Published:
  • 2014-10-13
Citation:
Tarabet, L., Lounici, M., Loubar, K., and Tazerout, M., "Dual Wiebe Function Prediction of Eucalyptus Biodiesel/Diesel Fuel Blends Combustion in Diesel Engine Applying Artificial Neural Network," SAE Technical Paper 2014-01-2555, 2014, https://doi.org/10.4271/2014-01-2555.
Pages:
8
Abstract:
Numerical simulation is a useful and a cost-effective tool for engine cycle prediction. In the present study, a dual Wiebe function is used to approximate the heat release rate in a DI, naturally aspirated diesel engine fuelled with eucalyptus biodiesel/diesel fuel blends and operated at various engine loads. This correlation is fitted to the experimental heat release rate at various operating conditions (fuel nature and engine load) using a least squares regression to find the unknown parameters. The main objective of this study is to propose a model to predict the Wiebe function parameters for more general operating conditions, not only those experimentally tested. For this purpose, an artificial neural network (ANN) is developed on the basis of the experimental data. Engine load and eucalyptus biodiesel/diesel fuel blend are the input layer, while the six parameters of the dual Wiebe function are the output layer. Levenberge-Marquardt (LM) learning algorithm is found to be the best learning algorithm with a minimum number of neurons in the hidden layer. The best results obtained by 2-12-6 network architecture show a good performance with a root mean square error (RMS) less than 0.009578, an absolute fraction variance (R2) in the range of 0.99825-0.99999 and a mean absolute percentage error (MAPE) in the range of 0.16-5.61%. Hence, the developed ANN model can effectively be used as a preferment prediction tool for the engine heat release rate.
Access
Now
SAE MOBILUS Subscriber? You may already have access.
Buy
Select
Price
List
Download
$27.00
Mail
$27.00
Members save up to 40% off list price.
Share
HTML for Linking to Page
Page URL

Related Items

Technical Paper / Journal Article
2010-09-28
Article
2016-12-11
Article
2017-03-13
Technical Paper / Journal Article
2010-10-19
Video
2017-03-26