Institute at the QTML 2024 in Melbourne
Hybrid quantum tensor networks: New approaches to stability prediction in aeroelasticity
At Quantum Techniques in Machine Learning (QTML) 2024 in Melbourne, Australia, the DLR Institute for AI Safety and Security presented a promising innovation: hybrid quantum tensor networks for precise analysis of complex physical systems such as aircraft wings or high-rise buildings exposed to wind.
Innovation through quantum AI
Stability prediction is a central component of aeroelasticity, a discipline that deals with the interaction between aerodynamic forces and elastic structures. Using a novel hybrid algorithm, DLR combines classical tensor networks with quantum technology. This approach makes it possible to analyse time series and provide early indications of stability problems.
A key advantage of the method is that local optimisation avoids the so-called barred plateau problem, a challenge in quantum computing that makes it difficult to train quantum models. This allows quantum computers to be used efficiently and reliably.
Applied research
The algorithm was tested on time series data from simulations carried out by the DLR Institute of Aeroelasticity. The results show that hybrid quantum tensor networks can not only predict the vibration behaviour of aircraft wings, but also offer potential for other physical systems, such as high-rise buildings or bridges under dynamic loads.
International networking and funding
Funded by ELEVATE and the Quantum Fellowship Programme, the project demonstrates how interdisciplinary collaboration is driving innovation in quantum AI. The results will be presented at the prestigious QTML 2024 conference, a platform for cutting-edge research in quantum methods and machine learning.
You can find out more about the projects and our research on our project pages.