Knock detection in spark ignition engines using a nonlinear wavelet transform of the engine cylinder head vibration signal

Abstract

This paper reports an investigation of knock detection in SI engines using a nonlinear wavelet transform based on the engine cylinder head vibration signals. The nonlinear wavelet transform employed in the current research is constructed by the redundant lifting scheme with an update-first framework, which has the ability to select an optimal wavelet function for each transform sample. The feasibility of applying a nonlinear wavelet transform to detect knock signatures from the vibration signals of the cylinder head is studied. The experimental results verify that the proposed method is capable of detecting knock signatures, even during the initial stages. More importantly, it is found that knock in each cylinder of a multi-cylinder gasoline engine can be detected using just one vibration sensor. Consequently, this is a simple and cost-effective way for knock detection in SI engines. Derived from the above method, we have developed a novel knock intensity factor and investigated the relevance of the proposed criterion for characterizing different levels of knock.