A real-time prediction interval correction method with an unscented Kalman filter for settlement monitoring of a power station dam.

A prediction interval (PI) method is developed to quantify the model uncertainty of embankment settlement prediction. Traditional PIs are constructed based on specific past period information and remain unchanged; hence, they neglect discrepancies between previous calculations and new monitoring data. In this paper, a real-time prediction interval correction method is proposed.

Physical interpretation of shear-rate behaviour of soils and geotechnical solution to the coefficient of start-up friction with low inertial number.

Shear experiments on soils have revealed the effects of shear rate, confining pressure, and grain size on the residual shear strength, but their nature is not well understood. To interpret these behaviours, a single dimensionless inertial number I from granular physics is introduced. A linear relationship between coefficient of residual friction μ and the natural logarithm of I was found by analysing geotechnical test data from other literature and helps to resolve the μ(aI)-rheology, which was proved invalid in the quasi-static regime.