English

Evaluating the phase dynamics of coupled oscillators via time-variant topological features

Data Analysis, Statistics and Probability 2021-03-24 v3 Machine Learning Algebraic Topology Chaotic Dynamics

Abstract

By characterizing the phase dynamics in coupled oscillators, we gain insights into the fundamental phenomena of complex systems. The collective dynamics in oscillatory systems are often described by order parameters, which are insufficient for identifying more specific behaviors. To improve this situation, we propose a topological approach that constructs the quantitative features describing the phase evolution of oscillators. Here, the phase data are mapped into a high-dimensional space at each time, and the topological features describing the shape of the data are subsequently extracted from the mapped points. These features are extended to time-variant topological features by adding the evolution time as an extra dimension in the topological feature space. The time-variant features provide crucial insights into the evolution of phase dynamics. Combining these features with the kernel method, we characterize the multi-clustered synchronized dynamics during the early evolution stages. Finally, we demonstrate that our method can qualitatively explain chimera states. The experimental results confirmed the superiority of our method over those based on order parameters, especially when the available data are limited to the early-stage dynamics.

Keywords

Cite

@article{arxiv.2005.03343,
  title  = {Evaluating the phase dynamics of coupled oscillators via time-variant topological features},
  author = {Kazuha Itabashi and Quoc Hoan Tran and Yoshihiko Hasegawa},
  journal= {arXiv preprint arXiv:2005.03343},
  year   = {2021}
}

Comments

13 pages, 8 figures

R2 v1 2026-06-23T15:22:37.858Z