Study On Settlement Of High Embankment On Analysis And Forecast And Settlement Control Standards

In recent years, highway construction is in a period of rapid expansion of development in China. More and more highway construction turned to the central and western regions. As the large central and western regions of elevations, there has been an increasing number of high fill embankment.There are more problems in High fill embankment stability and settlement control. The prevailing high fill embankment two technical problems: One is the construction of stability control, the other is settlement control after construction. These two issues require the settlement of High Embankment on a timely and accurate forecasts. Therefore, to make high fill embankment settlement research is necessary and meaningfulFirst of all, This paper describes the calculation and prediction of Settlement based on the current situation, from a high Embankment Settlement Mechanism, analyzes a variety of settlement prediction method of advantage and disadvantage.By Pearl growth curve model and the wavelet neural network theory analysis, because settlement of a single model has the time limitations and prediction accuracy is not high, combined the idea combination forecast, Paper presents an integrated growth curve model and neural network combination forecasting method of settlement.In this paper, the construction of buildings in Xiamen to Chengdu highway in Hunan section of Rucheng to Chenzhou highway to Subgrade Settlement and stability in the monitoring and information construction project. By wavelet transform time-frequency localization good nature and self-learning function of neural networks, combined with growth curve model, Thesis is established based on wavelet neural network combined forecasting model. Application of wavelet neural network combined forecasting model used to predict the settlement of High Embankment,The model was field measured settlement Ru-Chen highway data analysis and verification. The application of the model to avoid the calculation of various human factors, and growth curve model and BP neural network prediction results are compared, combined with high prediction accuracy, good agreement with the actual situation, with extensive project value and broad application prospect.Finally,Subgrade Settlement and stability control standard analysis. Thesis summed up by the roadbed construction and preloading of the settlement control standards. Then for settlement control after construction is no clear criteria to guide. Proposed car on the road driving comfort index as a control calculation of the settlement, Paper discusses the Settlement control standard after construction, with some practical engineering reference.