Parameter Identification of Time-varying Structure Research Based on Wavelet Transform
|School||Central South University|
|Course||Bridge and Tunnel Engineering|
|Keywords||Wavelet Transform Wavelet Ridge Instantaneous frequency Dynamic Programming Modal parameter identification Physical parameter identification Time-varying structure Nonlinear structure|
In recent years, more and more large-scale civil engineering structures installed structural health monitoring system, the structural parameter identification is a key core technology of bridge health monitoring and damage identification is difficult. For time-invariant systems, there are many modal parameter identification method. However, many civil engineering structures in the course of its operations showed varying characteristics, identify the time-varying parameters (modal and physical parameters) of this type of structure is very important for monitoring the structural operational status and diagnostic structural damage. In addition, the actual civil engineering structures exhibit a certain degree of nonlinear behavior nonlinear one hand, due to the nonlinearity of the material itself and the geometry of the structure, on the other hand due to structure damage also exhibit nonlinear behavior , the usual linear theory is not well reflect the essential characteristics of the non-linear structure. In this paper, on the basis of a summary of previous work using wavelet transform method to identify the time-varying variable structure parameters (modal and physical parameters) and the non-linear structure, from theory, numerical simulation and test three aspects more in-depth research. The main work is as follows: (1) briefly describes the current development of structural health monitoring system, pointed out the importance of structural parameter identification and nonlinear identification. Secondly, the linear time-invariant structure identification as well as the structure of the structural parameters of the time-varying nonlinear identification study conducted a detailed overview, clear research background, basis and meaning, to clarify the main contents of this paper. (2) the two extracted wavelet ridge method: an iterative method to extract phase information based on signal wavelet transform coefficients, wavelet ridge, the interference of the noise impact on the recognition result, based on the matrix singular value decomposition ( SVD) method to reduce the impact. Another method based on wavelet transform coefficient modulus information, according to the mode of great value, by applying a penalty function smoothing noise caused by the change of wavelet ridge discontinuity wavelet ridge extraction problem into optimization problems using dynamic planning method wavelet ridge, identifying the instantaneous frequency of the signal. (3) to establish a continuous complex Morlet wavelet transform wavelet ridge line identification structure instantaneous frequency-based approach, the design of a time-varying cable structure test. Suoshi Jia linear and sinusoidal variation of the tension, respectively, change the stiffness of the cable structure, the instantaneous frequency of the structure to achieve the time-varying, while the test structure of the impulse response, the use of the proposed method to successfully identify the instantaneous frequency of the cable, and the identification value structure inherent in the the fixed tensile force under modal frequency results were analyzed, the effect of the proposed method, the noise resistance. (4) change the structure of the physical parameters of the method based on the identification of discrete wavelet transform. Time-varying discrete wavelet transform variable structure parameters of multi-scale, expand the profile signal and detail signal, high-frequency detail signal is ignored, only varying parameters estimated by the low-frequency profile signal, and not too time-varying structure identification problem into variable structure identification problem. Then the least squares method to identify the low-frequency scale expansion coefficients to reconstruct the original time-varying parameters. The equation for the noise caused by ill-posedness of the problem, using Tikhonov regularization method to reduce its impact, and verified using numerical simulation and verification of the two-story frame structure with variable stiffness and damping. (5) The proposed method based on continuous wavelet transform to identify the structure of non-linear characteristics. Nonlinear structural skeleton curve concept of nonlinear structural systems approximate equivalent time-varying system, based on wavelet coefficients modulus maxima the ridge extracted wavelet transform and wavelet skeleton curve, identified the instantaneous frequency of the nonlinear structure and instantaneous amplitude, nonlinear structural skeleton curve, to identify the structure of the non-linear characteristics. Through with nonlinear stiffness structure numerical simulation to verify the effectiveness of the method.