A GREY TIME-SEQUENTIAL COMBINED MODEL FOR DYNAMIC PREDICTION OF HORIZONTAL DISPLACEMENT ON THE DEEP EXCAVATION WALL TOP

Horizontal displacement of the wall top of a deep excavation is a dynamic dependent process. During its excavation and construction, a equal intervals dynamic combined prediction model of recent information, in which grey system theory and time series analysis method is adopted, is established on the basis of measured data. With the addition of new deformation datum, the model parameters are continuously modified.The complicated factors influencing the displacement in the model are neglected. The practical displacement prediction of the pit has showed that the error between measure and prediction value is mostly below 5 percent.Under the condition of unstable displacement data, the combined prediction value is superior to that of the single grey model. But with the stable data, the prediction displacement value difference by combined or single model is little. The more prediction step, the more error.