Structural health monitoring and assessment using wavelet packet energy spectrum

Abstract

A novel methodology about structural health monitoring and assessment using the wavelet packet energy spectrum is developed by analyzing the acceleration response signal, aiming to identify the structural damage in real time and perform early structural damage alarming. The structural damage warning will be made according to baseline thresholds, which are obtained from a convergence analysis based on the mean and variance value of two structural damage indicators, namely energy ratio deviation (ED) and energy ratio variance (EV). To illustrate the effectiveness of the method, the operational stage of the Wangzong tunnel in the Wuhan Metro Line 3 in China is taken as a case study. The results demonstrate that (1) The multi-threshold wavelet packet denoising method is effective in removing noise and reserving information within 0.1 Hz; (2) The damage indicators ED and EV from the wavelet packet energy spectrum are sensitive to structural damage, whose mean and variance value will converge well along with the growth of wavelet packet component number; and (3) By the threshold value based on the one-side 98% upper confidence limit, the potential structural damage can be detected and alarmed dynamically in the underground filling with complexity and uncertainty. In conclusion, this research contributes to developing new metrics for structural health evaluation with consideration of time dimension and vibration properties of the structure, which is helpful in detecting and alarming the possible structural damage prior to structural failures. Its inherent limitations lie in the relatively complicated computing process and strict requirements in sensor location.