Wavelet package-neural network based on rough set diesel engine vibration signal identification model

摘要

To diagnose the wearing of diesel engine. Firstly, by applying the technology of wavelet package to analyze the quiver of the cylinder lid, five indexes of character information were brought. Secondly, the five indexes ones were reduced to the most important, decisive core index. At last, the core index was regarded as the input of the neural network. The output of the wearing modes of the valve were as normality, limitation, and leakage. Enough samples were used to discipline the neural network, so that the wearing state can be diagnosed correctly.