Study On Settlement Prediction Of High Speed Railway Tunnel Based On Improved Grey Time Series Combined Model

The appearance of high speed railway has promoted the development of national economy.In the high-speed rail construction,the construction of the high-speed railway tunnel is also in constant progress,and the construction of high-speed rail tunnel will be affected by various factors,in order to ensure the safety of construction process,deformation monitoring and prediction work has become very important.For the errors in the settlement observation data of Shen Nong tunnel,this paper uses the wavelet threshold denoising method to deal with the settlement observation data.Aiming at the disadvantages of the traditional gray model's background value parameter selection and the problems of the modeling samples.This paper optimizes the traditional GM(1,1)model based on the particle swarm optimization algorithm(PSO),and the background value of the model parameters based on Matlab prepared by iterative optimization procedures.At the same time,it improves the sample data of grey model to derive the equal dimensional GM(1,1)model,and optimizes the model parameters.In order to verify the fitting precision and the prediction precision of the improved model is better than before,use GM(1,1)model,PSO-GM(1,1)model and PSO-equal dimension GM(1,1)mode to analyze the settlement data,and the result is the fitting precision of PSO-equal dimension GM(1,1)mode is the best,and the best result is for XXmm.At the same time,the fitting residuals of the three models are analyzed by time series analysis,he combination of grey model and time series model are realized.The results of the three models are compared,and the results show that the prediction accuracy of the combined model of PSO-equal dimension GM(1,1)model and time series model are the highest,which can better predict the settlement trend.