Research On Temperature Effect Of Long-Span Cable-Stayed Bridge And Forecast Based On Neural Network

In recent years, technology of constructing cable-stayed bridge already entered the mature stage, but to the cable-stayed bridge in start construction and maintain support process to the ratio less of the influence research that the temperature variety result in, but the temperature effect is the more important factor to affect cable-stayed bridge. Therefore, the function that studying the variety of the temperature to rise to the cable-stayed bridge has the important meaning. This text launched the research at following works.On the foundation of developing circumstance of the detailed introduction domestic and international cable-stayed bridge in recent years, discuss the temperature variety to the output function of the cable-stayed bridge, and introduced the basic theories of the hot conduction, the formation that the temperature load, classification and characteristics, and domestic and international design standard in the provision that relevant temperature load.The study is based on Tie Luo Ping cable-stayed bridge, according to the Tie Luo Ping cable-stayed bridge designing, the data of construction and supervision reports, finite-elemented model of the bridge is established and the analysis is then carried out. Aim at a difference in temperature of season, the sunshine difference in temperature, reduceing suddenly difference in temperature these three kinds of different temperatures variety type suddenly to compute respectively, these the kinds of difference in temperature types from the various research to the influence that cable-stayed bridge produce, and analyze its main reasons.It usually has great affection in process of construction of cabled-stayed bridge with variety of the temperature, caused the main beam line form creation deviation in the end. In the problems of resolveing influence of cable force caused by temperature variety, put forward the lording method with technology of the Artificial Neural Networks(ANN), in introducing the basic theories of the Neural Networks and the foundation of the basic function and the estimating theories, introduce greatly Radial-Basis Function Neural Networks (RBF) to be the estimating network. This text be placed in the construction which in entry-level stage still and did not have the data actually because of the background engineering, so use the data constructed by the finite-elemented software to be the sample of the network estimating, see the data of difference from designed cable force in constructed and in designed temperature for the sample of importation, be used as the exportation sample to the changed quantity of cable force in construction, because of the non-linear mapping function of the Artificial Neural Networks, can predict by network which be trained well. By training, get the estimate result after computing, it is expressed that the predicting methods used in this text are reasonable and valid after the compareing of the results, it can provide dependable data in the aspects of avoid affecting in construction by the reason of temperature.In this text, the circumstances of the temperature changing are simulated and analyzed, use the technology of RBF Neural Networks in the works of predicting changed quantity of cable force in the affecting of changed temperature, acquired the satisfied result, express that the method is useful in the works which predict the cable force in the construction of cable-stayed bridge.