Journal
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
Volume 21, Issue 1, Pages 4-18Publisher
SAGE PUBLICATIONS LTD
DOI: 10.1177/1475921720904543
Keywords
Miter gates; artificial neural networks; surrogate model; finite element; inverse model
Funding
- United States Army Corps of Engineers through the US Army Engineer Research and Development Center Research Cooperative Agreement [W912HZ-17-2-0024]
Ask authors/readers for more resources
This article introduces a method of using Bayesian neural networks to learn damage features and using uncertainty estimates for cost-informed decision-making in structural health monitoring. An example of miter gates is presented to demonstrate the applicability of Bayesian neural networks.
Many physics-based and surrogate models used in structural health monitoring are affected by different sources of uncertainty such as model approximations and simplified assumptions. Optimal structural health monitoring and prognostics are only possible with uncertainty quantification that leads to an informed course of action. In this article, a Bayesian neural network using variational inference is applied to learn a damage feature from a high-fidelity finite element model. Bayesian neural networks can learn from small and noisy data sets and are more robust to overfitting than artificial neural networks, which make it very suitable for applications such as structural health monitoring. Also, uncertainty estimates obtained from a trained Bayesian neural network model are used to build a cost-informed decision-making process. To demonstrate the applicability of Bayesian neural networks, an example of this approach applied to miter gates is presented. In this example, a degradation model based on real inspection data is used to simulate the damage evolution.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available