Loading…
Effects of alcohol fuels on SACI engine and analyze of prediction of SACI engine performance by artificial neural networks
The combustion characteristics of spark assistant compression ignition mode under different loads and fuels were explored, as well as by constructing artificial neural networks(ANN) model, the prediction ability of different algorithms for engine performance was discussed. Burning methanol and n-but...
Saved in:
Published in: | Case studies in thermal engineering 2024-09, Vol.61, p.105040, Article 105040 |
---|---|
Main Authors: | , , , , , |
Format: | Article |
Language: | English |
Subjects: | |
Citations: | Items that this one cites |
Online Access: | Get full text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | The combustion characteristics of spark assistant compression ignition mode under different loads and fuels were explored, as well as by constructing artificial neural networks(ANN) model, the prediction ability of different algorithms for engine performance was discussed. Burning methanol and n-butanol can help to improve IMEP and induce earlier spontaneous combustion. The knock intensity(KI) of engine fueled with methanol was highest followed by n-butanol and gasoline, but maximum amplitude of filtered pressure oscillation(MAPO) shows the opposite trend. KI showed great positive liner correlation with ignition timing. Methanol showed the most outstanding tolerance on the compression ignition state. Fueled with methanol can decreased equivalent brake specific fuel consumption(ESFC) up to 52.9Â % compared with the initial SI gasoline engine. N-butanol can improve BSNOx, BSTHC and BSCO concurrently, however fueled with methanol will worsen BSNOx. ANN model for engine combustion, economy and emissions performance was built. The prediction accuracy of Bayesian regularization algorithm for engine performance predicting was highest, but it have no advantage in the calculating times when the amount of data was large while the Levenberg-Marquardt algorithm would be the ideal efficient. The mean square error of ESFC and emissions parameters under scaled conjugate gradients algorithm always poorest. |
---|---|
ISSN: | 2214-157X 2214-157X |
DOI: | 10.1016/j.csite.2024.105040 |