Wavelet Noise Reduction In Structural Damage Identification

Structural health monitoring has been widely studied in the field of civil engineering structures. Damage identification is an important part of structural health monitoring. The measured dynamic responses data, which includes both the useful information and the noise, is used in the vibration based damage identification. In consequence, the damage identification results is inevitably influenced by the noise data. Based on wavelet analysis, this thesis takes a deeper look at the influence of noise on damage identification, to eliminate the noise in the structure response signal and thus avoiding the interfere of damage identification from the noise.First, the structure response of common civil engineering structures is explored, by employing a various structural type, external extraction and so on. This work is helpful to understand the characteristics of structural response, and to distinguish it from the noise signal.Second, three wavelet denoising methods are used for signal denoising. The effect of denoising is evaluated by the signal to noise ratio and the root mean square of the error between original signal and processed signal. In addition, the selection of wavelet decomposition method is explored. By comparing the denoising results by various parameters selection, the wavelet noise reduction method, the mother wavelet choices, the decomposition level and the threshold selection rules are proposed. The proposed denoising strategy is verified by simply supported beam model for numerical simulation.Finally, structural damage identification is performed on a simple beam model, by various damage location, damage extend and noise level. The damage identification results is compared between the original data and denoising data. The effect of noise level to the damage identification results is studied to explore the effect of denoising method.Based on the wavelet analysis, this research studies the denoising methods and the selection of denoising parameters for the structural damage identification in the field of civil engineering. It is helpful for signal analysis of engineering structures.