A theoretical framework for Koopman analyses of fluid flows, part 2: from linear to nonlinear dynamics
Wei Zhang, Mingjun Wei

TL;DR
This paper develops a comprehensive theoretical framework for analyzing fluid flow dynamics using Koopman operator theory, linking linear and nonlinear behaviors through spectral analysis and mode decomposition techniques.
Contribution
It introduces a dual space approach to dynamics, revealing linear structures and extending Koopman analysis to nonlinear flows with infinite-dimensional eigenspaces.
Findings
Decomposition of flow past a cylinder using DMD.
Verification of Koopman-GSA equivalence at primary instability.
Identification of infinite Koopman modes in periodic flows.
Abstract
A theoretic framework for dynamics is obtained by transferring dynamics from state space to its dual space. As a result, the linear structure where dynamics are analytically decomposed to subcomponents and invariant subspaces decomposition based on local Koopman spectral theory are revealed. However, nonlinear dynamics are distinguished from the linear by local exponential dynamics and infinite dimension, where the latter is due to nonlinear interaction and characterized by recursively proliferated Koopman eigenspaces. The new framework provides foundations for dynamic analysis techniques such as global stability analysis (GSA) and dynamic mode decomposition (DMD) technique. Additionally, linear structure via Mercer eigenfunction decomposition derives the well-known proper-orthogonal decomposition (POD). A Hopf bifurcation process of flow past fixed cylinder is decomposed numerically by…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsModel Reduction and Neural Networks · Probabilistic and Robust Engineering Design · Fluid Dynamics and Turbulent Flows
