Random Matrix Ensembles in Hyperchaotic Classical Dissipative Dynamical Systems

TL;DR

This paper investigates the statistical properties of Lyapunov exponents in a dissipative, hyperchaotic classical system, revealing transitions from Poisson to GOE statistics and evidence of Tracy-Widom distribution in the largest exponent fluctuations.

Contribution

It demonstrates the emergence of universal random matrix statistics in the Lyapunov spectrum of a dissipative hyperchaotic system, linking chaos, dissipation, and random matrix theory.

Findings

Poisson statistics in nonchaotic overdamped limit

GOE statistics in hyperchaotic regime

Evidence of Tracy-Widom distribution in largest Lyapunov exponent fluctuations

Abstract

We study the statistical fluctuations of Lyapunov exponents in the discrete version of the non-integrable perturbed sine-Gordon equation, the dissipative ac+dc driven Frenkel-Kontorova model. Our analysis shows that the fluctuations of the exponent spacings in the strictly overdamped limit, which is nonchaotic, conforms to the \textit{uncorrelated} Poisson distribution. By studying the spatiotemporal dynamics we relate the emergence of the Poissonian statistics to Middleton's no-passing rule. Next, by scanning over the dc driving and particle mass we identify several parameter regions for which this one-dimensional model exhibits hyperchaotic behavior. Furthermore, in the hyperchaotic regime where roughly fifty percent of exponents are positive, the fluctuations exhibit features of the \textit{correlated} universal statistics of the Gaussian Orthogonal Ensemble (GOE). Due to the dissipative nature of the dynamics, we find that the match, between the Lyapunov spectrum statistics and the universal statistics of GOE, is not complete. Finally, we present evidence supporting the existence of the Tracy-Widom distribution in the fluctuation statistics of the largest Lyapunov exponent.