On using the wavelet transform to update models

Existing model updating techniques, based on modal data or frequency responses use experimental data to update the model of the structure with the implicit assumption that this data is stationary. However, noise present in the data is known to effect the computation of the updated parameters. In particular, the use of the Fourier transform averages the noise over the duration of the signal, smearing its effect over the frequency axis. One alternative to the Fourier transform is the wavelet transform, which has been successfully used in the electrical engineering and computer science fields to filter noise and compress data. This paper investigates two possible uses of the wavelet transform: in computing modal parameters and in formulating the model updating problem such that a formulation analogous to the Fourier-based approach is obtained. Numerical results show the feasibility of using the wavelet transform in updating models based on non-stationary or noisy measurements.