Condition detection of diesel engine cylinders based on continuous wavelet transform

To realize the non-disassembly detection of working diesel engine cylinders of armored vehicle, the misfire fault was simulated in an actual vehicle, noise signals nearby a vent in different states were measured. The low frequency component of periodical exhaust noise with obvious characteristics of amplitude demodulation was acquired by using continuous wavelet transform (CWT) under appropriate scale. Hilbert transform amplitude demodulating approach was used to extract the variance of the maximal amplitude normalized envelope signal as characteristic parameter, and a new method was presented for the working uniformity evaluation and misfire detection of diesel engine cylinders. The experimental results show that this method is effective and feasible.