Damage Identification Of Beam-type Structures Based On The Reconstructed Modal Rotations

The damage detection methods based on the structure vibration parameter often need to withdraw the structure from the actual structure vibration response signal the modality information. In structure online health monitor, the inconvenience of the certain vibration response quantity and the existence of the measurement noise enable the extraction of the modal parameter to have not the complete characteristic, and difficulty for the damage recognition. At present many research of the structure damage detection algorithm are about the shear pattern structural model, for it is difficulty to survey the structure corner information, and this leads to the barrier when algorithms applied to practical use. In order to solve this problem, we need to consider that the modality information restructuring (or is called status messages reconstruction). With this purpose, the main research works in this thesis are as follows:â‘ For the uniform stiffness distribution for beam structure, the static condensation method based on QR decomposition of the matrix is computed by the translational angle information, eliminating the physical parameters of structural materials, and export non-displacement parameters to be identified - rotation relationship by using MATLAB platform by modal translational movement angle displacement mode shape estimates.â‘¡By using soft threshold wavelet de-noising processing including noise, vibration signal, this paper discusses the characteristics of wavelet transform for signal, and analyzes the results of de-noising a great influence on the threshold identifying the problem.â‘¢By using the corner information reconstruction technology to build the improved cell damage variety D, the simple beam and continuous beam finite element simulation of the signal are exerted different levels of noise, and the degree of damage location and damage the effect of damage identification are investigated. The effectiveness of the information based on reconstruction of corner damage identification method is verified.â‘£Test the five rectangular beams separately. Firstly de-noising the acceleration response signal collected by wavelet soft-threshold to improve the signal to noise ratio; secondly stochastic subspace identification method using five beams extracted before and after injury in the first 3-order modal parameters, and finally curvature difference of modal flexibility and angle information based on reconstruction of the improvements are used to test cell damage variable beam detection, comparing the effect of detection before and after the de-noising. Further the feasibility and superiority of damage identification based on the information based on reconstruction of corner damage identification method applied to the actual structure are validated.