Applicability of the Bayesian Dynamic Linear Model Method in Dam Behaviour Analysis and Comparison with the HST Method

The detection of changes in structural behaviour over time is an important aspect in structural safety analysis. An anomaly detection method combining the existing Bayesian Dynamic Linear Models (BDLM) framework with the Switching Kalman Filter theory recently developed (implemented in OpenBDLM, an open-source software) is used on instrumentation data of Hydro-Québec dams. The key aspect of this method is its capacity to detect anomalies based on the prior probability of an anomaly, a generic anomaly model, as well as transition probabilities between a normal and an abnormal state. The approach is tested on different types of structural instrument data (pendulum, crackmeter, weir) and compared with results of the Hydrostatic-Season-Time (HST) method used in the field of dams.