Theoretical estimates for the largest Lyapunov exponent of many-particle systems

Abstract

The largest Lyapunov exponent of an ergodic Hamiltonian system is the rate of exponential growth of the norm of a typical vector in the tangent space. For an N -particle Hamiltonian system, with a smooth Hamiltonian of the type p 2 +V(q), the evolution of tangent vectors is governed by the Hessian matrix V of the potential. Ergodicity implies that the Lyapunov exponent is independent of initial conditions on the energy shell, which can then be chosen randomly according to the microcanonical distribution. In this way a stochastic process V(t) is defined, and the evolution equation for tangent vectors can now be seen as a stochastic differential equation. An equation for the evolution of the average squared norm of a tangent vector can be obtained using the standard theory in which the average propagator is written as a cummulant expansion. We show that if cummulants higher than the second one are discarded, the Lyapunov exponent can be obtained by diagonalizing a smalldimension matrix, which, in some cases, can be as small as 3×3. In all cases the matrix elements of the propagator are expressed in terms of correlation functions of the stochastic process. We discuss the connection between our approach and an alternative theory, the so-called geometric method.

Generalized Lyapunov exponents of the random harmonic oscillator: Cumulant expansion approach

“…This is essentially the approach followed by , Pettini et al [22][23][24], and the present authors [25][26][27]. Though there are some differences among the formulations of the three groups above, it may be said that the main theoretical conclusion extracted from that body of work is the following.…”

“…[25]). The comparison of theoretical results obtained with the cumulant expansion-truncated at the second order-versus numerical simulations has met mixed success.…”

An EMD-Based Chaotic Least Squares Support Vector Machine Hybrid Model for Annual Runoff Forecasting

“…Chaotic characteristics identification is significant to reveal the essential law of runoff series and establish a reliable forecasting model. The usual methods of chaotic characteristics identification include the phase portrait, power spectrum, saturated correlation dimension, largest Lyapunov exponent, Kolmogorov entropy, and so on [30]. The largest Lyapunov exponent is employed to identify the chaotic characteristics in this paper.…”

Statistically q‐deformed and tau‐deformed systems

“…Among these systems, one finds turbulent fluids [1,2], systems with electron-positron annihilation [3], systems exhibiting classical or quantum chaos [4,5], quantum entanglement [6], or long-range interacting many-body classical Hamiltonian systems [7]. Such systems, which are shortly classified as ''complex systems,'' are characterized by at least one of the following features: long-range interparticle interactions, long-term microscopic or mesoscopic memory, fractal nature of a pertinent subset of phase-space where the system remains long time or forever, small-size or scale-free systems (see [8] for a good introduction in the theory of complex systems).…”