Decentralized digital twins of complex dynamical systems
This addresses the problem of creating privacy-preserving digital twins for complex systems in computational science and engineering, representing an incremental advancement combining existing concepts.
The paper tackles the challenge of creating decentralized digital twins for complex dynamical systems by introducing a federated learning-based framework that enables collaborative model training without sharing raw data. The results demonstrate this approach can produce highly accurate digital twins for nonlinear spatiotemporal systems.
In this paper, we introduce a decentralized digital twin (DDT) framework for dynamical systems and discuss the prospects of the DDT modeling paradigm in computational science and engineering applications. The DDT approach is built on a federated learning concept, a branch of machine learning that encourages knowledge sharing without sharing the actual data. This approach enables clients to collaboratively learn an aggregated model while keeping all the training data on each client. We demonstrate the feasibility of the DDT framework with various dynamical systems, which are often considered prototypes for modeling complex transport phenomena in spatiotemporally extended systems. Our results indicate that federated machine learning might be a key enabler for designing highly accurate decentralized digital twins in complex nonlinear spatiotemporal systems.