Deep BBN Learning for Health Assessment toward Decision-Making on Structures under Uncertainties

Structural systems are often exposed to harsh environment, while these environmental factors in turn could degrade the system over time. Their health state and structural conditions are key for structural safety control and decision-making management. Although great efforts have been paid on this field, the high level of variability due to noise and other interferences, and the uncertainties associated with data collection, structural performance and in-service operational environments post great challenges in finding information to assist decision making. The machine learning techniques in recent years have been gaining increasing attentions due to their merits capturing information from statistical representation of events and thus enabling making decision. In this study, the deep Bayesian Belief Network Learning (DBBN) was used to extract structural information and probabilistically determine structural conditions. Different to conventional shallow learning that highly relies on the quality of the hand-crafted features, the deep learning is an end-to-end method to encode the information and interpret vast amount of data with minimizing or no features. A case study was conducted to address the methods for structure under viabilities and uncertainties due to operation, damage and noise interferences. Numerical results revealed that the deep learning exhibits considerably enhanced accuracy for structural diagnostics, as compared to the supervised shallow learning. With predetermined training set, the DBBN could accurately determine the structural health state in terms of damage level, which could dramatically help decision making for further structural retrofit or not. Note that the noise interference could contaminate the data representation and in turn increase the difficulty of the data mining, though the deep learning could reduce the impacts, as compared to conventional shallow learning techniques.