Single-trend component extraction for fault diagnosis of rotating machinery under time-varying speed conditions

Under actual operating conditions, vibration signals of rotating machinery often contain complex close-spaced components and strong background noise, which increases the difficulty of intrinsic chirp component decomposition (ICCD) to extract the fault characteristic components of rotating machinery. To tackle the above problem, a novel method, named single-trend component extraction (STCE), is developed in this article. First, a new decomposition framework is proposed by adopting a new penalty term designed in consideration of the low variation characteristic of instantaneous amplitudes to modify the optimization function of the ICCD, which improves the efficient distribution of energy between close-spaced components. Second, an instantaneous frequency (IF) estimation theory is proposed to obtain the IFs of the signal. Finally, a time-frequency representation with high energy concentration is obtained to reveal fault characteristic frequencies of rotating machinery. Both the simulation and experimental cases have confirmed the productiveness of the STCE in fault diagnosis of rotating machinery.