Skip to Main Content
Based on stationary wavelet packet transform and Hilbert transform, the paper proposes a fault detection algorithm, which adaptively extracts the fault characteristic component of the signal. Firstly, the algorithm uses one-level stationary wavelet packet transform to decompose the signal into low- and high-frequency sub-bands. Subsequently, Hilbert transform is used to obtain the instantaneous frequency and instantaneous amplitude of the low- or high-frequency sub-band. Based on the preset frequency and amplitude criteria, the algorithm decides whether to further decompose the sub-band or hold it. Thus the algorithm adaptively selects the path of stationary wavelet packet decomposition, making a multi-resolution spectral analysis on the signal and extracting the characteristic components for fault detection. The simulations show that the algorithm provides sufficient frequency-amplitude fault information with the less computational workloads and has better anti-noise performance.