Resumen
The problems of the synchroextracting transform method being unable to handle FM signals and being prone to time?frequency feature discontinuity in a strong noise environment are addressed by the construction of a novel rotation synchroextracting chirplet transform under the framework of the synchroextracting transform. The method retains the advantage of the generalized linear chirplet transform that can fit the time?frequency characteristics of the original signal and retains the high precision time?frequency analysis ability of the synchroextracting transform. The simulation results show that the proposed method is obviously superior to the generalized chirplet transform and synchroextracting transform method. The method can obtain the time?frequency energy located at the time?frequency ridges of FM-AM signals and multicomponent signals with crossed-frequency components, and has high time?frequency analysis ability and anti-interference ability. Finally, the proposed method is applied to diagnose mechanical faults. The experimental results further verify the effectiveness of the proposed method, which can effectively extract the characteristic freque.ncy of fault signal.