DeepMonitor
Deep physics based structural health monitoring
Principal Investigators
Prof. Dr.-Ing. habil. Stefan Kollmannsberger,
Chair of Computational Modeling and Simulation, Department of Civil and Environmental Engineering
Prof. Dr. rer. nat. Felix Dietrich,
Professor for Physics-enhanced Machine Learning, TUM School of Computation, Information and Technology
Project contributor
Divya Singh, M.Sc.,
Chair of Computational Modeling and Simulation, Department of Civil and Environmental Engineering
Qing Sun, M.Sc.,
Chair of Scientific Computing in Computer Science, Department of Informatics
Motivation and Goals
Structural flaws, especially when hidden, can compromise the integrity and safety of civil engineering constructions, such as buildings, bridges, and other infrastructure. Thus, non-invasive detection of hidden structural defects is a crucial task in civil engineering. The goal of the project is to develop a highly effective, data-driven framework that can detect structural defects in a computationally efficient manner. This framework will leverage advanced artificial intelligence (AI) and data science techniques to create a robust, efficient, and scalable solution for defect detection that can be used in practical civil engineering scenarios, enhancing structural safety and maintenance.
Recent Results Our newest approach focuses on two main aspects: (1) developing an efficient numerical solution method for the solution of the wave equation (forward solver) and (2) improving the non-destructive testing technique full waveform inversion (FWI) which fits computed to observed signals to reconstruct the material field .
- Wave equation solver (Fig 1):
- Neural network ansatz
- Mesh-free approach
- Spectral convergence observed in numerical studies
- Transfer learning based FWI (Fig 2):
- Physically meaningful minima
- Faster convergence
- Insensitive to initial weights of the Neural network
Selected Documentation
Journal papers | Herrmann, Bürchner, Dietrich, Kollmannsberger, "On the use of neural networks for full waveform inversion," in Computer Methods in Applied Mechanics and Engineering, vol. 415, pp. 116278, 2023. |
Conference papers | Bolager, Burak, Datar, Sun, Dietrich, "Sampling weights of deep neural networks," in Thirty-seventh Conference on Neural Information Processing Systems, 2023. Kollmannsberger, Singh, Herrmann, "Transfer Learning Enhanced Full Waveform Inversion*," in 2023 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), 2023, pp. 866-871. |