Application of innovation Kalman filter in tunnel deformation monitoring

Abstract

Due to the kalman filter can achieve good results while dealing with dynamic data, and the deformation data has dynamic characteristic, so in this paper, we use kalman filter to process data. In reality, the dynamic noise and observation noise are unknown previously, we often set initial values of noise based on the former experiences, and thus the precision of standard kalman filter will be effected by the inexact values of these two noise. However the innovation kalman filter could modify the dynamic noise and observation noise continuously, in this paper, this algorithm has obtained good results in tunnel deformation monitoring, and it has certain feasibility.