The analysis of the crankshaft speed fluctuation is one of the most investigated and used techniques for the detection and isolation of the misfire events in an internal combustion engine. During the past, lots of methods based on the time or frequency domain were presented in literature; this paper describes an analysis technique based on the wavelet approach that represents an efficient tool to analyze nonstationary signals. The use of a wavelet-based filter allows for the extraction of the frequency components related to a misfire event, and its localization in the time domain. The detection process is performed by analyzing the crankshaft angular velocity measurement, in order to isolate decelerations due to misfire events, and comparing to a dynamical threshold calculated through a Stationary Wavelet Packet analysis of the crankshaft velocity. Moreover, the proposed algorithm allows for an easy recognition of the cylinder responsible for the misfire. The paper presents experimental results supporting the validity of the described approach.
Misfire detection by a wavelet based analysis of crankshaft speed fluctuation / M. Montani; N. Speciale; N. Cavina. - STAMPA. - (2004), pp. 15-20. (Intervento presentato al convegno International Conference on Circuits, Signal and Systems(CSS 2004) tenutosi a Clearwater Beach, Florida, USA nel 11/28/2004 - 12/1/2004).
Misfire detection by a wavelet based analysis of crankshaft speed fluctuation
SPECIALE, NICOLO'ATTILIO;CAVINA, NICOLO'
2004
Abstract
The analysis of the crankshaft speed fluctuation is one of the most investigated and used techniques for the detection and isolation of the misfire events in an internal combustion engine. During the past, lots of methods based on the time or frequency domain were presented in literature; this paper describes an analysis technique based on the wavelet approach that represents an efficient tool to analyze nonstationary signals. The use of a wavelet-based filter allows for the extraction of the frequency components related to a misfire event, and its localization in the time domain. The detection process is performed by analyzing the crankshaft angular velocity measurement, in order to isolate decelerations due to misfire events, and comparing to a dynamical threshold calculated through a Stationary Wavelet Packet analysis of the crankshaft velocity. Moreover, the proposed algorithm allows for an easy recognition of the cylinder responsible for the misfire. The paper presents experimental results supporting the validity of the described approach.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
