Knock detection for gasoline engines based on the second generation wavelet transform

For precisely determining the detonation phenomenon of gasoline engines, the second generation wavelet transform was studied. Frequency aliasing is one of the three main distortions of the first generation discrete wavelet transform. A new type of the second generation wavelet transform was discussed to solve this problem. Compared with the first generation wavelet transform, the new type of the second generation wavelet transform is effective on suppressing frequency aliasing and saving computing time obviously. The new type of the second generation wavelet transform was then applied to the knock threshold detection of gasoline engine. The results show that the presented method can effectively extract the weak information of the knock from cylinder pressure signals and more complex signals of the cylinder head vibration. Compared with the first generation wavelet transform, the new type of the second generation wavelet transform is more suitable for the knock-based closed-loop ignition control for gasoline engines due to its faster computing speed.