Study On Structural Damage Identification And Threshold Value Classification In Structural Monitoring

The main work of the large bridge health monitoring system is to monitor the structural on line, identify the damage as well as the damage location and then evaluate the present state of the structure. The bridge damage identification and threshold value classifications which are difficult to study are a comprehensive system and involve many subjects such as bridge test specification, vibration theory, sensing technology, testing technology, system identification theory, signal analysis and processing technology, computer technology, evaluation theory and so on. But the obviously social, economic, technical significance and broad application prospect for the subject have been widely recognized. Therefore, the study of structural damage identification and the threshold value classifications have positive academic signification and practical value.After summarizing the methods of structural damage identification at home and abroad, some methods are proposed to study the structural damage identification. Considering the influence of the nonlinear structural vibration, the response signals of the nonlinear structure are analyzed by Hilbert-Huang transform. Finally, the study of threshold value classification of the structural monitoring is carried out. The main work and studies are as following:(1) The current developments of structural damage identification at home and abroad are reviewed. The methods of damage identification with modal information, time-frequency information and nonlinear vibration information are introduced respectively. The challenge and the difficulties in damage identification are summarized.(2) The structural nonlinear effects caused by breathing crack and the opening-closing effects of the crack are studied. The damage of the cantilever beam is identified based on nonlinear vibration. The breathing crack is simulated by contact elements in ANSYS. The acceleration signals of the free vibration and the forced vibration are respectively decomposed by empirical mode decomposition (EMD) and the energy ratios of the first two intrinsic mode functions (IMFs) of the signals are calculated. The results show that the breathing crack can cause high harmonics in acceleration responses and the effects of high harmonics are more obvious near the crack. For this reason, the approximate position of the crack can be located.(3) The method of damage identification is proposed based on singular value entropy theory. Taking the6-DOF spring-mass-damper system as an example, the singular value entropy is constructed with the IMFs of the displacements. The damage location can be identified according to the variations of the singular value entropy in different location before and after damage. The influences of the noise are discussed. The results show that this method has a good anti-noise property.(4) The damage identification method is proposed based on the kernel principal component analysis (KPCA). The kernel principal components of the frequency response function of the20-DOF spring-mass-damper system are extracted. The kernel principal components are used as the damage indexes. The damage location can be identified by the variations of the damage indexes before and after damage. In addition, the method shows better anti-noise property.(5) The method of threshold value classification with Fisher optimal division is proposed. As an example, the space finite element model of the Nanjing unite for the Nanjing Yangtze River Bridge is built. The displacements of the node at mid-span are calculated under different stiffness of the model. The displacements under different stiffness are constructed one sequence and then the displacement sequence is classified into3grades:the normal value, the abnormal value and the alarm value.