Congratulations to Charilaos Mylonas for successfully defending his PhD
Harry successfully defended his work on Machine Learning for Structural Health Assessment under Uncertainty with Applications in Wind Energy
In this doctoral thesis, Mr. Mylonas introduces tools from deep learning (DL) and scalable representation of uncertainty on deep neural networks to Structural Health Monitoring (SHM) problems, focusing on applications that are related to Structural Health Monitoring (SHM) for wind energy.
This work offers a general vision towards improving upon limitations of current SHM techniques. With focus on wind energy applications, it is proposed to achieve this vision through the combination of (1) flexible function approximation through deep learning, (2) scalable representation of uncertainty in deep neural networks, and (3) flexible incorporation of structure to data-driven algorithms using graph machine learning (GNs). It is argued that the limitations of current SHM techniques, both simulation-driven and data-driven, are related to computational and analytical simplifications, to which current state-of-the-art machine learning offers several improvements and a promising way forward.
Harry's work is summarized on a number of publications, found here:
https://chatzi.ibk.ethz.ch/about-us/people/person-detail.harry-mylonas.html