Machine learning based approach for forecasting emission parameters of mixed flow turbofan engine at high power modes

Hakan Aygun, Omer Osman Dursun and Suat Toraman

Energy, 2023, vol. 271, issue C

Abstract: To predict aircraft emissions from their own features has become more important as the usage field of aviation engines is extended to different sectors for different purposes. In the present study, thanks to by-pass ratio, pressure ratio and rated thrust of the one hundred sixty Mixed-Flow Turbofan (MFT) engines, NOx and CO emission indices as well as fuel flow for take-off and climb-out phases are predicted by using conventional Long-Short Term Memory (LSTM) at initial stage. After that, to improve the obtained estimations, hybrid Convolutional Neural Network (CNN-LSTM) model is employed to these engine dataset. The differences between two methods are presented by several types of error for each parameter. When considering the correlations, the relation between three variables and CO emission index is opposite whereas it is positively for NOx index. Moreover, NOx emission index of the MFT engines is predicted with 0.8166 of R2 by LSTM whereas its R2 increases to 0.8991 by means of hybrid CNN-LSTM approach. Amongst the models, the fuel flow is predicted the highest more than 0.95 of R2 whereas hybrid model makes it predictable with the higher accuracy, which has more than 0.99 of R2 for both phases. Moreover, these enhancements are observed in predicting of the other parameters regarding MFTs. It is thought that the proposed CNN-LSTM model could help in estimation of important parameters of gas turbine engines with higher level of accuracy.

Keywords: Turbofan emissions; Long-short term memory; Convolutional neural network; Machine learning; Gas turbine engine (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (7)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544223004206
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:271:y:2023:i:c:s0360544223004206

DOI: 10.1016/j.energy.2023.127026

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

More articles in Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().