Research on application of GM(1, 1) model based on Kalman filtering in monitoring of Zhengzhou-Xi'an Dedicated Passenger Line

The Kalman filtering and the GM(1, 1) model are widely applied in the engineering deformation situation analysis, but most of the application is based on single model. The GM(1, 1) model can be easily affected due to the stochastic disturbance in the modeling process, but the greatest advantage of Kalman filtering is that the stochastic interference noise can be excluded. In this paper, through the application practice in monitoring of Zhengzhou-Xi'an Dedicated Passenger Line in the collapsible loess area, it is proved that GM(1, 1) based on the Kalman filtering can better reflect the deformation trend of the observed object to a certain extent, and has certain advantages in the application in the engineering deformation monitoring, the analysis results can serve as the reliable theoretical basis for the post-construction settlement and deformation prediction, and a new research idea is provided for the post-construction deformation monitoring and deformation prediction of the high-grade highway and railway in the collapsible loess area.