Extended wavelet transforms in acoustic diagnosis

We propose an efficient strategy for qualitative acoustic diagnosis of electric motors based on fast wavelet transforms and fuzzy set theory. First an extended wavelet transform is applied to the acoustic data recorded by a microphone. Thus a decomposition of the frequencies in the data sets is obtained and only a few typical features (wavelet coefficients) remain for the identification process. Then a fuzzy c-means algorithm separates data into various clusters thus giving a qualitative interpretation of the present fault due to the calculated membership values.