On Hilbert-Huang Transform Approach for Structural Health Monitoring

摘要

The Hilbert-Huang Transform (HHT) approach is used for signal processing to discover the possible damages existing in a structure. A program is written in Fortran and in Matlab, and then verified by two simple cases with known solutions. The HHT method is then utilized to analyze the actual benchmark data of UBC (University of British Columbia) for the purposes of structural health monitoring. The results demonstrate that HHT method has the capability of: (a) recovering the actual signals' time-frequency feature; (b) detecting and locating the actual structural damages; (c) detecting the instant of the impact load or of the damage taking place in active structural health monitoring. The method of Empirical Mode Decomposition (EMD) is used to eliminate noise. Results show that EMD method is better than the Wavelet method for the de-noising purpose.