Time Series Prediction of Shield Tunnelling Posture
In order to solve the problem of difficulty in controlling the posture of the shield machine in the process of tunnelling,this paper builds an autoregressive integrated moving average(MF-ARIMA)based on median filter to predict the shield posture of the shield machine in the process of tunnelling,rel-ying on the engineering example of the Geng—Qing section shield tunnel project of Hefei Metro Line 7.The research results show that MF-ARIMA has a good performance in predicting the shield posture,with the prediction coefficients of determination of the horizontal and vertical deviations of the front shield are 0.81 and 0.93 respectively,and the prediction coefficients of determination of the horizontal and horizon-tal deviation of the shield tail are 0.86 and 0.94 respectively.The average prediction coefficients of de-termination of MF-ARIMA is 0.89.The MF-ARIMA model is used to predict the shield posture,predict the risk of the shield posture exceeding the limit in advance,and combine with the on-site engineering measures to reduce the cementing time of synchronous grouting,which can effectively reduce the shield posture deviation during tunnelling and improve the shield construction quality.
NETL
NSTL
万方数据
