The Model Updating Of Wuhu Yuanze Bridge Based On Environmental Excitation

Bridges in service are inevitably affected by environmental erosion,material aging,repeated load and other factors.Therefore,the loading behavior of bridges in services is changed.There is a big difference between the finite element model established by the original design drawings and the existing bridge,so the initial finite element model needs to be updated to the current state.In this paper,a concrete filled steel tubular arch bridge is taken as an example.The dynamic data was obtained by health monitoring system.Then the operational modal parameters are identified by peak-picking(PP)method and stochastic subspace identification(SSI)method.Through the sensitivity analysis of the parameters to the natural frequency,the main parameters to be updated are selected.Accordingly,the initial finite element model is updated by the first-order optimization algorithm and the BP neural network response surface method,and the advantages and disadvantages of the two methods are compared.The outcomes of this thesis can be a reference for other bridges finite element model updating.The main work and conclusions are as follows:1.The initial finite element model is established based on the design drawings.The dynamic characteristics of the structure are calculated,and the natural frequencies and vibration modes are obtained.2.The Yuanze bridge health monitoring system in Wuhu city,Anhui Province,is described.Then the modal analysis is carried out according to the ambient vibration data obtained from the health monitoring system.The natural frequencies of the bridge are identified by PP method and SSI method.Incorporated both PP and SSI methods can guarantee identified results.The identified modal parameters can be used as the basis for the updating of the finite element model.3.One of the key issues of finite element model updating is how to choose the corresponding updated parameters.The five parameters that have a large influence on the natural frequencies are selected as the corresponding parameters by using sensitive analysis method.4.The calculated frequencies of initial finite element model are compared with those of measured frequencies.The initial finite element model of Yuanze bridge is updated by the first order optimization algorithm and BP neural network response surface method.The results show that the two methods are consistent.Therefore,both the first order optimization algorithm and the BP neural network response surface method can be used to complete the finite element model updating.But two methods have their own advantages,should be properly selected.