Wavelet transform-based modal parameter identification considering uncertainty

This paper presents an integrated identification method to consider the uncertainty effect on modal parameters for output-only system. The method is based on the time-frequency characteristics of the wavelet transform (WT) and the capabilities of the bootstrap distribution in statistical estimation. For the WT-only based identification method, the important issues related to identification accuracy such as modal separation, end-effect, associated with the parameter selection of wavelet function based on Shannon entropy, are given detailed investigations. The bootstrap procedure is then employed to evaluate the uncertainty effects by providing confidence interval of the modal parameter statistically. The effectiveness of the integrated method has been confirmed through numerical simulation and experimental test on a bridge model.