ABSTRACT
The gearbox, a vital element of any plant machinery, always needs special attention as it undergoes significant environmental conditions throughout its service life. The tracking of gear performance degradation is paramount to ensure the reliability of the whole system. This performance degradation is generally based on the vibration-based condition monitoring program, which entails statistical health indicator trending. However, traditional health indicators are not sensitive to incipient faults at a very early stage because the actual fault signature is generally masked under environmental noise. Hence, this chapter proposes a signal denoising methodology based on maximal overlap discrete wavelet transform (MODWT) to obtain a health indicator with early warning capability for incipient fault and exhibits a monotone trend. First, the raw vibration signature is denoised with the help of a state-of-the-art MODWT signal processing technique to clearly identify the hidden fault signatures. Then various traditional statistical features are extracted from this denoised signal. These multidimensional features are then processed with principal component analysis (PCA) to establish a single-dimensional health indicator to quantify the fault severity. Performance comparison of the proposed method with traditional analysis is presented and the proposed method outperforms at every level.
