Harmonic-Gaussian double-well potential stochastic resonance with its application to enhance weak fault characteristics of machinery

Noise would give rise to incorrect filtering frequency-band selection in signal filtering-based methods including fast kurtogram, teager energy operators and wavelet packet transform filters meanwhile result of useful components even mode mixing, end effects, etc., decomposition-based empirical decomposition, singular value decomposition local mean decomposition. On the contrary, noise stochastic resonance (SR) is beneficial enhance weak signals interest embedded with strong background noise. Taking into account that nonlinear systems are crucial ingredients activate SR, here we investigate SR cases overdamped underdamped harmonic-Gaussian double-well potential subjected a periodic signal. We derive measure analytic expression output signal-to-noise ratio (SNR) steady-state probability density (SPD) function under approximate adiabatic conditions. When loses its stability, can observe antiresonance phenomenon, whereas adding damped factor system change stability potential, resulting behavior disappears system. Then, use characteristics for diagnosing incipient rotating machinery failures. Theoretical experimental results show adjusting both intensity parameters which there bell-shaped peak SNR. Additionally, independent frequency-shifted rescaling process large machine parameter outperforms one. Finally, comparing advanced robust (RLMD) method based on cancellation or infogram filtering, characterize better performance detect Fault early stage failures successful improving fault characteristic identification rolling element bearings.