Developing an integrated structural health monitoring and damage prognosis (SHM-DP) framework for predicting the fatigue life of adhesively-bonded composite joints

Early damage assessment in large-scale structures by innovative statistical pattern recognition methods based on time series modeling and novelty detection

“…In this regard, one of the most efficient unsupervised learning methods is novelty detection [43]. Although novelty detection methods may provide some limitations in detecting damage type and damage prognosis [44], they are widely used in SHM applications. In general, a novelty detection strategy for SHM is subdivided into a baseline (training) phase and an inspection (monitoring) phase.…”

A Bayesian Approach to Model Calibration and Parameter Estimation in Potential Drop Measuring

“…Bayesian inference has been widely used in model-based structural health monitoring and prognostics in the context of damage diagnosis and damage prognosis, where sensor data of any kind is used in conjunction with physics and measuring characterising models to assess the state of a single component or a whole structure while taking into account system-inherent uncertainties and variabilities, e.g. [11,12]. In the presence of unknown parameters that govern the physics-based models, for example parameters of a damage evolution model, the inverse problem becomes even more complex.…”