Modal Parameter Identification And Finite Element Model Updating Of Large Span Suspension Bridge Based On SHMS

Modal parameter identification and finite element model (FEM) updating of the bridge structure are both main research hotspots in the field of structural health monitoring of the large span bridge. Accurate identification of structural modal parameters and FEM updating can lay the foundation for status evaluation, damage detection and relative maintenance of the bridge structure. And they can also provide references for reasonable design, construction and reparation of the similar bridge. In this thesis, the Runyang Suspension Bridge (RSB) is taken as engineering background. All the works are firmly focused on the two active topics, modal parameter identification and FEM updating of large span suspension bridges. On the basis of wind data and structural vibration data recorded during typhoons by the bridge Structural Health Monitoring System (SHMS), the research on the modal parameter identification and FEM updating of large span suspension bridge are conducted. The main works are as follows.(1) Analysis on the non-stationary wind characteristics of the measured typhoons at the bridge site of RSB. The wind velocity and direction of typhoon "Matsa" and "Khanun" recorded by the SHMS of RSB are taken as research objects. The wind characteristics, such as mean wind speed, turbulence intensity, turbulence integral scale, turbulence power spectral density, are calculated based on the stationary wind speed model and non-stationary wind speed model, respectively. And the results are compared with each other. Research results show that the measured wind speed during typhoons exhibits obviously non-stationarity and the non-stationary wind speed model is more suitable.(2) Comparison research on modal parameter identification of large span suspension bridge based on three kinds of typical methods. Based on the vibration data of RSB recorded by its SHMS, modal parameters of its main girder are identified by Averaged Normalized Power Spectral Density (ANPSD), Stochastic Subspace Identification (SSI), Hilbert-Huang Transform (HHT), respectively. And the results are compared with each other, which aims at determining the applicability of the three methods to identify modal parameters of RSB. Results show that measured acceleration responses during typhoons exhibits obviously non-stationarity and the identified modal parameters based on the non-stationary-applicative HHT method are more accurate in this situation. Therefore, the HHT method is chosen as the main modal parameter identification method in the later research.(3) Study on relationship model between modal parameters of large span suspension bridge and its monitoring environmental factors. Modal parameters of the main girder of RSB and the corresponding wind velocity, temperature during typhoon are acquired. And these data are pre-processed by Nonlinear Principal Component Analysis (NLPCA). On the basis of this, the Artificial Neural Network (ANN) model between environmental principal component and modal parameters is established. Then, the distribution pattern of fitting residuals between the measured modal parameters and ANN-based predicting values is analyzed so as to examine its accuracy. Finally, the environmental-factor-elimination modal parameters are obtained from the ANN-based nonlinear predicting model. The research results can lay the foundation for the FEM updating of RSB.(4) FEM updating of large span suspension bridge based on Relative Vector Machine (RVM)-Markov Chain Monte Carlo (MCMC). On the basis of initial finite element model of RSB, Latin Hypercube Sampling method is used to obtain test samples, which can be applied to build the dynamic response surface model. With the combination of structural modal parameter measured from SHMS and field test, the finite element model is updated based on MCMC. Then, the updated modal parameters are compared with the initial and measured parameters and it aims at confirming the reliability of the model updating method. The research results can provide reference for FEM updating and damage detection of other similar large span bridges.