Back to Search Start Over

The Use of Multicriteria Inference Method to Identify and Classify Selected Combustion Engine Malfunctions Based on Vehicle Structure Vibrations

Authors :
Krzysztof Prażnowski
Andrzej Bieniek
Jarosław Mamala
Adam Deptuła
Source :
Sensors, Vol 21, Iss 7, p 2470 (2021)
Publication Year :
2021
Publisher :
MDPI AG, 2021.

Abstract

Internal combustion engines are among the most commonly used propulsion units for drive systems in various industries such as land transportation, maritime transportation, and power generation. Their operation involves a continuous change of technical condition as a result of not only the combustion process but also their operation under conditions of variable load or ambient impact. It is therefore important to monitor the technical condition of internal combustion engines to ensure high performance and reliability over their lifetime. The article presents the test results obtained from incorrect operation of an internal combustion engine as a result of forced failures of the ignition and injection system. On this basis, a multicriteria comparison of the signal analysis of engine block vibrations was made, after the transformation of the signal from the time domain to the frequency domain, by using the induction technique obtained from the operation of decision tree algorithms. For this purpose, the amplitude spectrum in the frequency domain, scaled to absolute values of discretization for which teaching and testing data tables were created for successive harmonics, was determined for the engine block vibration signal being tested. On the basis of the developed algorithm using decision trees, a multicriteria data table was analyzed for which a compatibility path for the analyzed engine block vibration signal is created. In this way, it is confirmed with a specified degree of effectiveness, depending on the point of operation of the engine resulting from its crankshaft speed, that there is a possibility of detecting a preset ignition or injection system malfunction in the technical condition of the internal combustion engine with a probability up to about 72%.

Details

Language :
English
ISSN :
14248220
Volume :
21
Issue :
7
Database :
Directory of Open Access Journals
Journal :
Sensors
Publication Type :
Academic Journal
Accession number :
edsdoj.9eb91ab98f045459a576f8e2feca47d
Document Type :
article
Full Text :
https://doi.org/10.3390/s21072470