Research Of Monitoring And Prediction Method For Settlement On The Bridge Of Ballastless High-speed Railway

Safety of high-speed railway construction is concerned with the speed of high-speed railway construction is being accelerated and the export of high-speed railway construction technology. But in the process of building high-speed railway, the settlement monitoring and forecasting of the foundation is particularly important throughout the whole building phase and the operation phase. During the pre-construction, settlement monitoring is needed to control the safety of the construction; In the medium-term, both the settlement monitoring and forecasting are needed to determine the reasonable time of track laying. In the operational phase,the settlement monitoring and forecasting is also necessary to ensure traffic safety. The basis foundation of high speed-railway are mostly constructed with viaduct, so the study of settlement monitoring and forecasting of ballastless track viaduct in the construction of high-speed railway is becoming a top priority.Currently, the conventional methods of settlement monitoring and forecasting of high-speed railway includes curve fitting method and system for forecasting model. The curve fitting methods mostly include hyperbolic method、three-point method、Asaoka and logarithmic method; And the GM(1,1) model of Gray theory is common in the system for forecasting model. In this paper, these four common methods will be selected to analyze.At the same time, this paper studies the effect of the wavelet denoising to settlement prediction ability. First of all, throughout the contrast test, select the better wavelet basis function which has better denosing ability, and then make a comparison of the settlement prediction using the original data and denoising data. The test concludes that settlement prediction with the wavelet denoising data can effectively improve the prediction accuracy and improve the stability of prediction.By selecting the foundation settlement observation data of the basis foundation in the second section of Tianjin to Qinhuangdao Passenger and on the basis of the existing overall predictive analysis throughout the construction phase, then denoise the settlement data, and then the whole construction is divided into three phases. Firstly, from the pier poured to the girder’s placement. Secondly, from the girder’s placement to the track laying; The last phase is from the track laying to trial operation. In these stages, the four methods are used to analyze and The conclusions below is made:In the first phase, the GM(1,1) model is best and the three-point method is better; In the second phase,the hyperbolic method is best, GM(1,1) model and the three-point method are better; in the last phase,the GM(1,1) model and the hyperbolic method are both good, but taking into account the actual situation of the project and the GM(1,1) model requires the same temporal distance data, the hyperbolic method is the best in the third phase.