Parameterized Iterative Time–Frequency-Multisqueezing Transform for Bearing Fault Diagnosis

As a core component of rotating machinery in industry and manufacturing, rolling bearings are affected by a variety of dynamic changes, which brings challenges to the monitoring and fault diagnosis of bearing operation status. Therefore, a new parameterized iterative time-frequency multisqueezing transform (GTFMST) method is proposed to improve the fast time-varying signal time-frequency resolution for bearing fault diagnosis. Firstly, Fourier spectrum is employed to find the optimal parameter coefficients of the kernel function in generalized Warblet transform(GWT). Secondly, a new parameter transformation kernel estimation method based on the Fourier spectrum is proposed to solve the problem that the GWT of the signal suppresses the time-frequency energy diffusion. Thirdly, a new multiple iterative rearrangement strategy in both time and frequency directions is designed to achieve higher time-frequency energy concentration. Finally, the numerical simulation signal and real-world fault signal are employed to prove the effectiveness of the proposed method. The experiment results demonstrate that the GTFMST exhibits superior time-frequency energy concentration, more effectively captures the time-frequency attributes of the signals and achieve more precise diagnostic outcomes.