Comparative analysis of different engine operating parameters for on-board fuel octane number classification

• Main Authors: Ghanaati, A., Muhamad Said, M. F., Darus, I. Z. M.
• Format: Article
• Published: Elsevier Ltd 2017
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://eprints.utm.my/id/eprint/75356/; https://www.scopus.com/inward/record.uri?eid=2-s2.0-85020838300&doi=10.1016%2fj.applthermaleng.2017.06.013&partnerID=40&md5=d15f3af1a18c0d93eedf7e352643098d
• Institution: Universiti Teknologi Malaysia

The comparative analysis on combustion of commercial gasoline with research octane number (RON) 95, 97, and 100 was carried out on a spark ignition (SI) engine under different engine speeds, loads and spark advances. The RON classification procedure was investigated using regression analysis and artificial neural network (ANN) by executing the combustion properties derived from the in-cylinder pressure signal and engine rotational speed signal. The results showed a special pattern for each fuel RON; these patterns were obtained using the peak in-cylinder pressure, maximum rate of pressure rise, and maximum amplitude of pressure oscillations. In addition, a pre-defined threshold or formula is necessary to restrict the implementation of these parameters for on-board fuel identification. Lastly, the confusion matrix that provided the ANN model efficiency for RON classification had the highest accuracy when the pressure signal was employed as the network input for all spark advance timing. However, the ANN model with rotational speed signal input could only identify the fuel octane number after a specific advance timing that was detected at the beginning of noisy combustion because of knock. The confusion matrix for the ANN with speed signal input increased from 68.1% to 100% when ignition advanced from −10° to −30° before top dead center. The results established the feasibility of use of the rotational speed signal as the input for an ANN model to identify different fuel octane number patterns.