| The bridge structure is always located in the throat of the traffic network line,which is related to people’s life safety and the development of national economy.Therefore,it is of great significance to install the health monitoring system on the bridge to realize the function of damage warning and health evaluation.The realization of the function of health monitoring system depends on the accurate finite element model.It is the only way to obtain the required precision model by establishing the refined initial finite element model with the model modification technology.Firstly,the basic theories of bridge modeling,modal parameter identification and model modification are introduced.The modeling strategies commonly used in long-span Bridges are discussed,and the fine modeling methods for continuous girder Bridges,cable-stayed Bridges and arch Bridges are introduced in detail.The selection principle of modified parameters,optimization algorithm and criterion of correlation in the modification of parametric model also discussed.Secondly,the basic principle of stochastic subspace method is introduced in detail.A structural modeling method by combining top-down modeling method and direct modeling method is proposed.This method can effectively improve the efficiency of model revision,and the more complex the structural model is,the more obvious the improvement of the modeling efficiency is.This paper introduces the principle of zero order optimization algorithm,discusses the process of model modification using this algorithm,and completes the development of model modification program based on zero order optimization algorithm on ANSYS platform.Finally,the modeling method and model modification method mentioned above are applied to three actual engineering Bridges.The results shows that the stochastic subspace method can effectively identify the modal parameters of the actual structure,and has the advantages of strong anti-noise,clear concept and stable identification results.It is suitable for modal parameter identification under environmental excitation.The proposed structure modeling method can effectively reduce the time of fine modeling.Combined with the zero-order optimization algorithm,the model was modified to effectively improve the coincidence degree between the model and the actual structure.The modified model reached the expected goal and was suitable for the health monitoring system. |
