2021 Volume 77 Issue 2 Pages I_477-I_484
In this study, targeting a consolidation problem, we combine a data assimilation method based on time series observations with reliability analysis to update the conditional limit state probabilities. The settlement is predicted by using soil/water coupling FEM analysis, and its uncertainty is estimated by using a particle filter as a data assimilation method. Three types of hypothetical time hitories of settlement are prepared as observation data with different magnitudes of settlement based on forward analysis of the soil/water coupling FEM. The limit state probabilities at observed and unobserved locations are updated based on the increase of observation in time. It is shown that the predicted settlement and the conditional limit state probability are updated quantitatively according to the increase of observation data, which leads to resonable decision-making.