Study On The Construction Control Of Long-span Prestressed Concrete Continuous Bridge

With the continuous development of our traffic, a large number of bridges need to be constructed. Prestressed concrete continuous bridge is favored because it has fine performance of the force, small deformation, comfortable during driving, beautiful styling and strong seismic capacity. Prestressed concrete continuous bridge always constructed by cantilever casting technology. During the construction, its structural system has been a state of flux and the stress and displacement of the structure changes. In order to ensure the quality and safety of the bridge in construction stage, bridge construction control is essential. This paper has conducted in-depth research of prestressed concrete continuous bridge construction control on the background of Xisubao BridgeBy learning existing methods of construction control, for comparison, this paper selected adaptive control method as the control method of this bridge. Generally, there are three bridge control calculation method. They are forward-calculation method、back-running method and stress-free status method. By studying and comparing their advantages and disadvantages, the forward-calculation method was chose. In this paper, the Xisubao Bridge was simulated by using the software MIDAS/CIVIL. It can review the design of the bridge and provide theoretical data for the later construction control. According to the calculation results, during the construction phase, the maximum stress is12.5MPa^22.68MPa; during the operational stage, the maximum stress is11.2MPa≤16.2MPa. They meet standard requirements. At the same time, the simulation results also provide theoretical data for the construction control. In this paper, structural design sensitivity parameters were analyzed. From the analysis, the deflection is sensitive to the change of deadweight and the section inertia moment, and the section stress is sensitive to the change of deadweight and prestress. The structure, algorithm derivation and achieving with MATLAB language of BP Neural Networks were introduced. The problems existed in BP Neural Networks were analyzed. By studying commonly used methods for improving algorithm, a improved algorithm was used to train the networks and forecast. This also lay the theoretical foundation for the practical application of the BP network.The construction control of Xisubao Bridge was implemented. The construction control system and construction control processes were created in site. Real-time monitoring of the state of the bridge was implemented. By collecting, processing and analyzing the data, the result in construction stage is that the maximum stress is11.22MPa≤22.68MPa, meeting standard requirements. The error between theory data and measured data is less the20%. The bridge is in the safe and controllable state. The linear control of this bridge was implemented. Forecasting the elevation by using the BP Neural Networks. According to the results of the sensitivity analysis, the main parameters that affected the elevation forecast can be determined. Training the improved BP Neural Networks by using the data collected in the early stage. Forecasting the elevation by using the trained BP Neural Networks. Comparing with the measured results, the error of forecasting results were less than20mm. It can meet the construction requirements. The error of shut section is19mm≤22mm. It can meet the standard requirements. The bridge closed successfully.