Inferring coupling strength from the Kuramoto order parameter

Citations

WEB OF SCIENCE

0
Citations

SCOPUS

0

초록

Accurately estimating the coupling strength in oscillator networks from macroscopic observations is essential for predicting synchronization transitions. We consider the inverse problem of reconstructing the unknown coupling strength K in the globally coupled Kuramoto model from scalar observations of the macroscopic order parameter R(t), assuming that the natural frequencies and the initial phase configuration are known. After initialization, individual phase trajectories are treated as hidden, and only the scalar order parameter is observed. We employ an extended Kalman filter with an augmented state representation that recursively estimates the coupling strength from observations of R(t). By exploiting the mean-field structure of the globally coupled Kuramoto model, the covariance prediction step can be computed efficiently, substantially reducing the computational cost. The numerical simulations demonstrate that the proposed estimator accurately reconstructs the coupling strength and remains stable even when R(t) is small and strongly fluctuating. © The Korean Physical Society 2026.

키워드

Extended Kalman filter; Inverse problem; Kuramoto model; Parameter estimation; SYNCHRONIZATION; POPULATIONS
제목
Inferring coupling strength from the Kuramoto order parameter
저자
Kim, Gug Young; Sul, Hoseok; Choi, Jee Woong; Son, Seung-Woo
DOI
10.1007/s40042-026-01739-w
발행일
2026-10
유형
Article in press
저널명
Journal of the Korean Physical Society