Continuous wavelet grey moment approach for vibration analysis of rotating machinery

Abstract The statistical characteristics of the continuous wavelet transform scalogram of vibration signal are explored, and two features, wavelet grey moment (WGM) and first-order wavelet grey moment vector (WGMV), are proposed for condition monitoring of rotating machinery. Faulty signals from a rotor test rig are studied using the first-order WGM. The analysis indicates that first-order WGM can reveal the time–frequency features of vibration signals well and that it could be used to diagnose faults quantitatively. The effectiveness of the first-order WGMV is also demonstrated by experimental data. Results show that the first-order WGMV is suitable to reflect the local information of scalogram, and would be an effective method of vibration signal analysis for fault diagnostics of rotating machinery. The result by FFT analysis is also given. Compared with the conventional FFT method, the two features presented in this paper are more suitable to extract characters of faults quantitatively.