The Settlement Calculation And Research On Settlement Prediction Of The Deep Artificial Fillroadbed

Due to the influence of the natural environment and geographical conditions,in the western part of China,especially in Yunnan and Guizhou areas,the terrain is undulating and complicated,leading to deep artificial fill roadbed is very common,and settlement is a key issue affecting its engineering quality.There are many factors influencing the settlement of the roadbed in the actual engineering,and the variation law,interrelations and the degree of cognition of the settlement laws are often difficult to analyze clearly.There are some limitations in calculation of the settlement and prediction,so it is of great theoretical and engineering significance to accurately calculate and predict the settlement of theroadbed.On the basis of Petro China Yunnan petrochemical field handling special railway line deep artificial filling station project,combining on large amounts of settlement observation data,making a study on the settlement calculation and settlement prediction of deep artificial fill with the experimental study and theoretical analysis,this thesis obtains the following conclusions:(1)This paper analyzes the applicability of different settlement calculation methods to the deep artificial fill roadbed,based on the Petro China Yunnan petrochemical field handling special railway line deep artificial filling station project.According to the actual situation of the scene,choosing a typical work site for drilling and sampling,considering the characteristics of artificial filling,taking the basic physical and mechanical properties test and single compression consolidation test,obtaining the physical parameters,compression coefficient and compression index of artificial fill to take the calculation theory for settlement and comparative analysis the applicability of different settlement calculation methods to the settlement calculation of deep artificial fill.The result shows that modified formula method and the stress history method have certain guiding significance for the final settlement of artificial fill foundation.(2)According to the prediction error of settlement data uncertainty factors under the settlement,the recurrence plot and recurrence quantification analysis method of settlement time series forecasting were analyzed,to identify the settlement time series good prediction effect.Through the predictability analysis of the settling time series of deep artificial fill roadbed: The settlement time series is chaotic;The more regular of the color distribution and the more uniform in the shade,smaller white areas appear along the main diagonal,and the predictability of settlement time series is stronger.Quantitatively analyses the recurrence plots on this basis,extracting the recursive quantitative index curve changes over time and immediate determining the starting point of prediction time is 22 days.Using the recurrence plots to get the second order renyi entropy as the predictable time length of the settlement time series to comprehensive extrapolates the best settlement time of the forecast effect is C4.(3)The settlement of the deep artificial fill roadbed is predicted by the unbiased gray prediction model with residual Markov chain correction.The fitting curve is consistent with the measured curve,which verifies the accuracy of the model.In the actual monitoring process,the collected settling time series is non-equidistant.Therefore,the project firstly used three cubic spline interpolation function of settlement time series time interval,then used the residual Markov chain to revise the unbiased grey forecasting model to predict the settlement data,The predicted results are in good agreement with the experimental results which show that the model is verified.