Sensitivity Analysis And Identification Of Parameters On Construction Control Of Short-Line Method

With the development of transportation, more and more longspan bridges are built upon the river and the fjord. And the wide use of PC continuous beam bridge accelerates the construction technique. In China, the construction technique of cantilever erection based on short-line method is applied and spread under this background. As the time which been spent on applying short-line method is limited, the time to use Construction control of short-line method is comparatively short in China as a result. The theory of closed-loop control is adopted in the control process to alignment control. That is to say, the construction process is simulated before the construction, and the alignments of prefabricatin,erection and completion are calculated. Then take the first two alignments as the goal in the prefabrication and erection process, so as to make sure that the alignment of the completion will be realized.This thesis aims to make a contribution to the development of Construction control of short-line method from the theory of closed-loop control to the theory of self-adaptive control by sensitivity analysis and identification of parameters. In the process of sensitivity analysis of parameters of the C4 continuous bridge of the JiMei Bridge. in this thesis, the method of sensitivity analysis of parameters based on BP Neural Network is emphasized and compared with classical method. at mean time, it is proved that this method is applicable in Construction control of short-line method, and the method of sensitivity analysis of parameters in construction control of bridge is enriched.In this thesis, four main parameters, dead weight of segment,rigidity of segment,preservation time of segment and tensioning force, are educed according to sensitivity analysis of parameters. BP neural network is used to identify the four main parameters synchronously, and the applicability of BP neural network in identification of parameters on construction control of short-line method is proved theoretically.