On a quantitative method to analyse dynamical and measurement noise
Authors:
- M. SiefertMSM. Siefert
- J. PeinkeJPJ. Peinke
- R. FriedrichRFR. Friedrich
Preprint
Abstract
This letter reports on a new method to analyse empirical time series of a dynamical system with respect to the types of noise involved. Based on experimental and numerically generated data we show that it is possible to differentiate between dynamical and measurement noise. This method is based on the theory of diffusion processes, it neither depends on further assumptions, as the form of the process, nor on any parameters.Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- Experimental indications for Markov properties of small scale turbulenceProceedings in Applied Mathematics and Mechanics, 2002
- Extracting model equations from experimental dataPhysics Letters A, 2000
- Estimation of Noise Levels for Models of Chaotic Dynamical SystemsPhysical Review Letters, 2000
- Delay embedding in the presence of dynamical noiseDynamics and Stability of Systems, 1998
- Takens embedding theorems for forced and stochastic systemsNonlinear Analysis, 1997
- Resource Letter: ND-1: Nonlinear DynamicsAmerican Journal of Physics, 1997
- Noise reduction in chaotic time-series data: A survey of common methodsPhysical Review E, 1993
- THE RECONSTRUCTION OF THE FOKKER-PLANCK AND MASTER EQUATIONS ON THE BASIS OF EXPERIMENTAL DATA: H-THEOREM AND S-THEOREMInternational Journal of Bifurcation and Chaos, 1993
- Unbiased estimate of forces from measured correlation functions, including the case of strong multiplicative noiseAnnalen der Physik, 1992
- Multimode oscillations in a modified Van Der Pol oscillator containing a positive nonlinear conductanceProceedings of the IEEE, 1981