The Control Of Linear In Long-span Continuous Beam Bridge By Cantilever Casting

With the rapid development of highway construction in our country, more and more long-span Bridges have been built, which promoted the rapid development of long-span prestressed concrete continuous girder bridge. The construction of large span continuous beam bridge is a complex process. The cantilever construction method as a continuous beam bridge construction method more and more is used of continuous beam bridge during cantilever construction needs to undergo structural system conversion, is bound to make complex changes in the displacement of girder. The elevation with the beams of construction completion is very difficult to adjust, only through early prediction and follow-up of vertical mold elevation beams to adjust. It is necessary to monitor the whole process in the process of construction In order to ensure the quality of bridge construction and security, can make the project go smoothly, also to make when it was built to design the structure of the linear of hope and reasonable internal force state.Put forward the construction control scheme of Chong He Bridge and combine with the red crane bridge engineering practice, to identify the impact of parameters for generating configuration state deviation. By adopting the method of parameter sensitivity analysis, Is the influence parameters of main girder structure in the construction process and into a state of bridge were analyzed, and the influence of the main influence parameters, using the least squares method to main influence parameters identification and correction. We established the analysis model by the finite element analysis software Midas/Civil and carries on the analysis of main girder construction phases. Determine girder deformation value at every stage of the theory, and the theoretical calculated values and measured values were compared analysis, determine the correctness of the theoretical model and the results. Application of neural network to forecast the main girder alignment. In view of the traditional BP neural network slow convergence speed and easy to fall into local minimum point of problem, put forward the corresponding improvement measures, and build a model to forecast the main girder alignment with the help of neural network toolbox in MATLAB7.0. According to the final girder elevation predicted values and the measured results, concludes that the improved BP neural network used in the linear control of cantilever construction of continuous girder bridge is feasible.