Parameter Sensitivity Analysis In Construction Control Of Long Span Hybrid Beam Cable-Stayed Bridges

Two parts about construction control of long span hybrid beam cable-stayed bridges are included in this thesis: parameters sensitivity analysis and parameters identification based on BP neural network.The finite element model of Edong Yangtze River Bridge is set up according to the real construction stages of it. The parameter sensitivity analysis is studied by the calculation, which is carried out by changing the structure parameters.The influence of the bridge behavior, because of the change of the structure parameters, is simply explained with theory and through that the correctness of the result is verified.Parameters, which may remarkably affect the calculation results of bridge behavior, are determined by parameter sensitivity analysis and identified by neural network technology. The feasibility of the application of the neural network method in the construction control of long span hybrid beam cable-stayed bridge is explored. First, by analyzing the Edong Yangtze River Bridge, using the nonlinear analysis system-NLABS, samples of learn,train and test are obtained. Then, a parameter identification model of construction control is established based on BP neural network. A parameter identification program is designed by using MATLAB programming language. At last, the application is verified feasible in the construction control of long span hybrid beam cable-stayed bridge by comparing its result with the real value of parameters which are identified.