Parameter Identification Of Construction Process Of Cable-Stayed Bridge Based On Bayesian Method

With the rapid development of China's transportation industry,the field of bridge construction is also making great strides.Various long-span PC cable-stayed bridges have been widely used.The cable-stayed bridge is a high-order statically indeterminate structure.The influence of construction error on the structure cannot be ignored,and the actual state after the bridge is changed.The construction error is cumulative.If the deviations in the construction process are neglected,the default structural parameters are always consistent with the design values and have been in the theoretical state.Then,as the construction progresses,the final actual bridge state is likely to be different from the design state.Unification leads to misjudgment of the actual state of the structure.This will have an impact on health monitoring,maintenance and reinforcement work after the bridge.Because these tasks are based on the real state of the current structure,not the theoretical state of the design.Therefore,it is extremely necessary to use the measured data and some processing methods to update the key parameters of the structure in real time and correct the finite element model in time to make the output theoretical value more accurate and minimize the error.It is said that the Bayesian method in mathematical statistics is introduced,and the structural parameters of PC cable-stayed bridge in real time are updated,and the finite element model is corrected,and the measured results to verify whether the Bayesian method is applicable are compared with.At the same time,the least square method commonly used in the field of parameter identification is used to compare the recognition results of the two to verify the accuracy and efficiency of the Bayesian method.In this way,the bridge can be obtained in a more realistic state,which is an important basis for future management,maintenance,repair and reinforcement.At the same time,it also adds a new and effective method for the field of bridge parameter identification.The main conclusions of this paper are as follows:(1)A new parameter identification method is introduced,Bayesian method,its basic content is detailed,and how to apply it to the parameter identification work in the field of bridge engineering is clarified.(2)A finite element model based on the engineering structure is established.Taking the measured deflection and cable force as the target response value during the construction process,the Bayesian method is used to continuously update and update the model,and the theoretical value after correction is compared with the measured data to verify the parameter identification.(3)At the same time,in this case,the least square method which is more popular in the field of parameter identification is used to identify the structural parameters.The results of the two methods are compared and analyzed in detail.The final results show that the two methods are basically in the same precision,and the Bayesian method has a better overall recognition effect.(4)It shows that there is a difference between the actual state and the design state of the bridge operation stage,and this difference is recognized to a certain extent.