PhD Student

• Scientific machine learning
  • Physics-informed neural networks
  • Combination of numerical methods and machine learning
• Hybrid digital twins of critical infrastructures
  • Model- and data-based digital twins

Possibilities of using physically informed neural networks for critical infrastructures. (Working Title)

Critical infrastructure structures are of particular engineering importance, especially with regard to their stability and monitoring. The stability of such structures is usually ensured by physical models and standardized verification procedures. Structures can be monitored using so-called digital twins, which account for changes in boundary conditions and actions during the life of a critical infrastructure structure and provide an indication of the extent to which mechanical strength and stability may be compromised. As a result, structures are inspected or maintained at more frequent intervals to address potential failure.

Recent findings in the field of physically informed neural networks have made it possible to combine the physical modeling that forms the basis of digital twins with the potential of machine learning.

The goal of this work is to investigate the potential of this linkage with respect to classical engineering problems and to connect to the current state of research. In this context, open questions in the fundamentals as well as potential applications in the field of critical infrastructure are addressed.