Rotor test platform vibration state detection based on improved harmonic wavelet

Harmonic wavelet analysis can extracts the singular components effectively in the non-stationary signals. Firstly, this paper elaborates on the structure and decomposition of the harmonic wavelet, and it also proposes an improved harmonic wavelet algorithm which based on the improvement of the harmonic wavelet frequency domain features and the change of the time domain signal instantaneous cutting features, it makes the excellent frequency domain features of harmonic wavelet being more serious on its time localization properties. Secondly, this paper also introduces a kind of capacitive vibration acceleration sensors briefly which is used to get the vibration signals of rotor test platform. Mapping out and explaining the time-frequency diagrams based on the improved harmonic wavelet algorithm. Finally, the experiment simulated a set of vibration signals with singular points to express the rotor test platform vibratory fault state. The experiment proves that the improved harmonic wavelet analysis method can implement the rotor test platform vibration state detection by analyzing the abnormal signals, and its effect is better than traditional Fourier analysis method.