The Application Of BP Neural Network On Cable-stayed Bridge's Construction Control

With the fast development of the transportation business and the unremittingly technique of bridge construction, bridge construction to across the degree, highly difficult direction development greatly, structure supple, beauty and safety of cable-stayed bridge gradually drive extensive adoption. Construction control is an important part of cable-stayed bridge construction technique; it is a key to make sure the construction quality of bridge. So it is an important work to cable-stayed bridge's construction control. Factors that affect construction control many and complex, and these information independent each other. How will valid exploitation is these information, combine modern mathematics theories, build up related supervision model and carry on a valid construction control, is the main contents of this text research. The full text combined with the characteristics of large cable-stayed bridge construction, research the artificial neural network used to construction control, works mainly did as follows:(1) With the analytical and discussion of the basic principle to the artificial neural network, from theories, the article expounds and proved the possibility of the neural network method used to construction control. In this foundation top, the factors that influence cable-stayed bridge's construction control and its interaction relation are researched.(2) At the foundation of analytical of the blemish that BP study calculates way existent in physically applied, give a homologous improvement method. Combine an analytical result to establish a model that in keeping with the BP artificial neural network, to predict the formed gearing line control of the steel box of cabled-bridge.(3) Aim at a greater problem of the construction data unit and the quantity class difference, the adoption a method of data return on turning, and overcame the class bad data which influence of the astringency disadvantage to the network from the self-study.(4) Combine the example the engineering of the third Nanjing Yangtze River Bridge, drawled up the line form control prediction procedure with the MATLAB procedure language, obtained a good result.(5) Compare the gray theories prediction and the neural network method prediction; proved the artificial neural network method an application to has superiority more in large cabled-bridge construction control.