Structural damage identification using modified Hilbert–Huang transform and support vector machine

In the current study, a new structural damage detection algorithm is presented using the modified Hilbert–Huang transform and support vector machine. The modified Hilbert–Huang transform is an adaptive time–frequency analysis tool that alleviates the mode mixing issue encountered with Hilbert–Huang transform. On the other hand, since the measured vibration responses are generally nonlinear and non-stationary signals, the Fourier transform utilizing the sinusoidal functions is inadequate for their processing. Thus, the modified Hilbert–Huang transform is utilized to study the measured signals. The structural damage features are constructed with the Hilbert spectrum energy of selected intrinsic mode function obtained by decomposing the measured vibration signals with modified ensemble empirical mode decomposition. The support vector machine’s classification and regression algorithms are utilized to detect the location and extent of the damage, respectively. The offshore platform's experiment model is utilized for theoretical and experimental validation of the presented method's effectiveness.