Translated Title
Technology of Voice Recognition Based on Wavelet Transform
Translated Abstract
Information of the acoustic signal is abundant, collection of it is easy, so people can use signal processing method to analyse objects＇＇ sound by computer and extract its feature.These features can offer reliable prediction for object＇＇s running.The paper introduces the process of voice recognition, including denoising of the acoustic signal, separation the prominent parts of the acoustic signal.Experiment results show that it is an effective and reasonable method to diagnose fault of gas engine comparing with other fault diagnoisis methods.
Translated Keyword
acoustic signal
denoising
mechanical failure
the layered threshold
