Investigations On Structure Damage Identification Based On Hilbert-Huang Transform

Structural health monitoring has become the hot spot in the academic and engineering both at home and abroad, has a strong vitality and research value. Damage identification technology is the most key link in health monitoring, so the research of structure damage identification has important theoretical and practical significance. A review of present research on structural damage identification technology is presented in this thesis as well as its development trend. Briefly introduces several kinds of commonly used time-frequency analysis method and points out their shortcomings, which raises a new signal processing method:Hilbert-Huang Transform (HHT).HHT is a new technology for the analysis of the nonlinear and non-stationary signals. This method consists of two successive parts, i.e., the empirical mode decomposition (EMD) and the Hilbert spectral analysis:firstly, an arbitrary signal is decomposed into a number of intrinsic mode functions (IMF) by EMD; then, Hilbert transform is performed on each IMF, and the corresponding instantaneous frequency is obtained, as well as its Hilbert spectrum and energy spectrum of signal. In this method, the main conceptual innovations are the introduction of intrinsic mode functions based on local properties of the signal, which makes the instantaneous frequency more physical meaningful.Centering on the HHT, this paper systematically introduces the basic concepts and theoretical basis of this method. Summarizes the shifting stoppage criteria selection problem and boundary effect of the EMD sieving process, and proposes the corresponding handling method. In order to demonstrate the effectiveness of the method, the author used this method to analysis a synthetic signal.Through an example of modal mixing effect, this paper raises the ensemble empirical mode decomposed (EEMD) which could alleviate the modal mixing problem. In order to solve the problem of adding noise level of EEMD, noise-polluted data are projected into phase space by time delays and added noise levels in EEMD are determined through the singular values in the process of signal phase space reconstruction. In this paper, author presents numerical simulate signal to evaluate the effectiveness of this approach. Furthermore, also presents experimental validation of this approach using a transmission tower model to obtain its vibratory response signal and fundamental noise level. During the study, the author had found the proportion noise. And a noise model of damage identification is put forward. The author points out than the noise of measured signal include base noise and external noise. The based noise is only related to the test system; the external noise is related to the amplitude of test signal, increases as the test signal amplitude increases.The HHT method was applied to structural damage identification, and a damage index is presented, which based on EMD energy. In this article, the effectiveness of the EMD energy damage index for damage detection is demonstrated through a set of numerical and experimental investigations. In the numerical study, a three-story frame structure with quality damage or stiffness damage, located at different locations along the structure. In the experimental investigation, which used the same transmission tower model as the noise experiment to obtain its acceleration response signal, different damage case of the tower were examined. In both the numerical and experimental studies proved the capability of the EMD energy damage index for detection and quantification of damage degree and therefore can be regarded as an effective tool for structural damage identification purposes.