Fuel Recognition Using Particle Swarm Optimization of Classification Trees

Abstract:

Work cycle control of a compression ignition internal combustion engine based on cylinder pressure and injection pipe pressure allows optimizing many aspects of engine performance. Selected descriptors of pressure courses can be used for constructing models and classifiers to describe engine work conditions and fuel supply. The work presents the methodology for obtaining classifiers that recognize the type of fuel injected into a cylinder with accuracy acceptable in practical technical applications. These classifiers were constructed by means of two different methods from the field of computational intelligence: classification trees induction and particle swarm optimization of the classification trees. The accuracy and transparency of the proposed models were compared.

• Periodical: Applied Mechanics and Materials (Volume 712)
• Pages: 23-28
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.712.23
• Online since: January 2015
• Authors: Leszek Radziszewski, Michał Kekez*
• Keywords: Computational Intelligence, Fuel Recognition, Particle Swarm Optimization (PSO)

References

[1] L. Radziszewski, M. Kekez, Classification of fuel injected into cylinder of internal combustion engine by computational intelligence methods, in: R. Dwornicka et al., Modeling of dynamic systems, Kraków, 2014, pp.42-61. (in Polish).

[2] L.S.G. Teixeira et al., Multivariate calibration in Fourier transform infrared spectrometry as a tool to detect adulterations in Brazilian gasoline, Fuel 87 (2008) 346–352.

DOI: 10.1016/j.fuel.2007.05.016

[3] M. Kekez, L. Radziszewski, Modelling of pressure in the injection pipe of a diesel engine by computational intelligence. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering (2011) 0954407011411388.

DOI: 10.1177/0954407011411388

[4] R. Isermann, N. Müller, Design of computer controlled combustion engines, Mechatronics (2003) 1067-1089.

DOI: 10.1016/s0957-4158(03)00043-6

[5] M. Yoon et al., A method for combustion phasing control using cylinder pressure measurement in a CRDI diesel engine, Mechatronics (2007) 469-479.

DOI: 10.1016/j.mechatronics.2007.06.001

[6] J. Pietraszek, A. Gądek-Moszczak, N. Radek, The estimation of accuracy for the neural network approximation in the case of sintered metal properties, Studies in Computational Intelligence 513 (2014) 125-134.

DOI: 10.1007/978-3-319-01787-7_12

[7] J. Pietraszek, N. Radek, K. Bartkowiak, Advanced statistical refinement of surface layer's discretization in the case of electro-spark deposited carbide-ceramic coatings modified by a laser beam, Solid State Phenomena 197 (2013) 198-202.

DOI: 10.4028/www.scientific.net/ssp.197.198

[8] J.R. Quinlan, Data mining tools See5 and C5. 0, http: /www. rulequest. com/see5-info. html.

[9] J.R. Quinlan, C4. 5: programs for machine learning, Morgan Kaufmann, (1993).

[10] M. Kekez, L. Radziszewski, Fuel recognition in compression ignition engine in the real time, Pomiary Automatyka Kontrola 59 (2013) 470-472.

[11] A.P. Engelbrecht, Fundamentals of computational swarm intelligence, John Wiley & Sons, (2006).

[12] Information on http: /www. particleswarm. info/Programs. html.