Displacement Model for Concrete Dam Safety Monitoring via Gaussian Process Regression Considering Extreme Air Temperature

Abstract

Structural health monitoring models provide important information for safety control of large dams. The main challenge in developing an accurate dam behavior prediction model lies in the modeling of extreme temperature effect. This paper presents a Gaussian process regression-based displacement model for health monitoring of concrete gravity dams, which can model the temperature effect by using long-term air temperature data. Important attractions of Gaussian processes include accurate simulation results, convenient training, and so forth. Different covariance functions and temperature variable sets are tested on the horizontal displacement prediction problem of concrete dams. Results show that segmented air temperature based Gaussian process regression models can reflect the extreme air temperature effect on displacements of concrete gravity dams, considering the prediction accuracy is much better than that of a mathematical model based on periodic functions.