Study On Sensitivity Of Elevation Control Of Prestressed Concrete Single Tower Cable Stayed Bridge

This paper takes the elevation control of the prestressed concrete cable-stayed bridge as the background and the parameter sensitivity analysis as the research foundation,considering the influence of various parameters on the linear state of the main girder during the construction process,and realizes the linear control of the main girder during the construction process.Firstly,through parameter sensitivity analysis,the importance relationship of different response results based on the sensitivity parameters of the engineering bridge is obtained.Determine the main girder weight error and stay cable initial tension deviation are the two most sensitive parameters that affect the bridge structure response,and further explore the influence of the main girder weight error and the stay cable initial tension deviation on the main girder linear change based on this law.Then the calculation results of the main beam weight and the initial tension of the stay cable varying with the segments are discussed.The variation of the parameters of the block itself and the linear changes at its own cross-section are explored,and the linear changes of the main cross-sections of the other segment parameters are also obtained.The law of change.It is found that the mid-span area is the most sensitive area affected by parameter variation in each section of the main girder,and it is also the area that deserves the most attention for linear control.At the same time,using the law of mutual influence between the blocks,the envelope diagram of the weight error control of the concrete segments during the construction process is obtained.Finally,a parameter error recognition method is proposed based on the dendritic neural algorithm.This method is based on the Taylor series expansion principle,by constructing the target polynomial,using the computer’s data processing advantages for sample training,outputting the coefficients of the polynomial term,and establishing the mapping relationship between the input and output physical quantities.As long as the displacement data in the actual project is collected,the parameter error value that needs to be identified can be deduced backward.At the same time,based on the algorithm’s excellent filtering ability for white noise,as long as there is a unified relationship between the physical quantities selected as neurons,they can be used to replace the value of the original physical quantity and output the same result.Compared with the traditional least square error recognition method,the dendritic neural algorithm has the ability to recognize white noise,which cannot be realized by the least square method.