Research On Moving Load Identification On Bridge Based On Function Approximation Method

In recent years, moving load identification from the bridge's response has aroused an universal interest in academia. After summarizing the existing research results, this article proposes new methods on moving force identification based on radial basis function(RBF) approximation method and cubic spline function approximation method, and discusses the influence which some parameters to the precision of identification through calculating an example.In first chapter, the article first introduces some existing methods which are used to identify moving load and the research situation both in home and abroad, and proposes in this foundation a goal and its significance, namely to propose new moving load identification methods and discuss the precision of new methods through numerical simulation, so that we can assess bridges' dynamic performance, and provide the basis for the bridge design and the automobile performance improvement. At the same time, this chapter also cleares about the research work which this article must complete.The second chapter introduces the basic principle and the correlation knowledge of moving load identification, including mode shape orthogonality, mode superposition principle and so on.The third chapter and the fourth chapter are the key parts of whole paper, which separately applies RBF approximation method and cubic spline function approximation method in moving load identification on a bridge. The third chapter makes the simple introduction to the nerve network, and proposes the research mentality and the steps of moving load identification. RBF network has a strong approaching ability, applying it in moving load identification can obtain precise results. The influence of some parameters is discussed, such as the mode number, the load moving speed and the measured point number and their position.The fourth chapter discusses the reason why the cubic spline functions could be used on this topic, cubic spline functions has a good smoothness and can ideally fit the dynamic response curve of a bridge. Its algorithm is simple, and has preciserresults than RBF approximation method. It also discusses the influence of the parameters above. The results indicate that the parameters have strong impact on all the two method, so we should carefully select these parameters.Finally, the fifth chapter summarizes the work in this article, and it's proved that so long as the parameters are suitably choused, all the two methods above can effectively identify the moving load on bridge. Also, it pointes out the insufficiency of these two methods, and views the next step of the research work.