Prediction control for construction of long-span continuous girder bridge based on improved grey GM (1,1) model

摘要

The alignment monitoring of the large-span cantilever bridge is intractable in construction. High-accuracy prediction is required in order to control the construction accurately. Thus, the grey control theory is widely applied to this field. But existing GM (1,1) prediction model is little equipped with measured data, which undermines its accuracy. In response to drawbacks of GM(1,1) prediction model, this paper subjects the original sequence to smooth processing using the whole data model, puts in place an adjusted grey system control program and applies the program to the alignment monitoring of large-span beam-arch composite bridge which is a (60+128+60)m inter-city expressway running from Beijing to Tianjin. The comparison results of measured deformation and predicted deformation show that the improved grey control system has fairly high prediction accuracy.