Noise stabilization can occur in nonlinear dynamical systems driven by noise. The essential idea is that the noise-free system is unstable, but with noise present the system exhibits bounded persistent fluctuations. An extreme example would be a dynamical system that has no bounded behavior without noise (e.g. almost all orbits eventually reach infinity), but which exhibits bounded persistent behavior for any finite level of noise. Such systems are qualitatively different in character from deterministic chaotic systems, yet they can mimic low-dimensional chaos, implying that new time series methods must be developed to distinguish noise-stabilized systems (where the attractor is `created' by the noise) from a `noisy' strange attractor (where the noise simply `fuzzes' out the natural probability density of the noise-free deterministic system). In this talk we will discuss a simple dynamical model, motivated by a problem encountered in plasma physics, that exhibits this behavior. After a brief summary of the theoretical background and motivation, we will describe what we believe to be the first experimental realization of this phenomenon: a nonlinear circuit that is noise stabilized.
展开▼
