Gear tooth diagnosis using wavelet multi-resolution analysis enhanced by Kaiser's windowing

Wavelet multi-resolution analysis shows promising results for gear tooth damage diagnostics. However, selecting an accurate mother wavelet, defining dynamic threshold value and identifying the resolution levels to be considered in gear fault detection and diagnosis are still challenging tasks. This paper proposes an enhanced wavelet-based technique for detecting, locating and estimating the severity of defects in gear tooth fracture. The proposed technique improves the wavelet multi-resolution analysis by decomposing the noisy data into different resolution levels with data sliding through Kaiser's window. Only the maximum expansion coefficients at each resolution level are used in de-noising, detecting and measuring the severity of the defects. A small set of coefficients is used in the monitoring process without assigning threshold values or performing signal reconstruction. The proposed technique shows accurate results in detecting and localizing gear tooth fracture.