Time series analysis for system identification and diagnostics (Q2638980)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Time series analysis for system identification and diagnostics |
scientific article |
Statements
Time series analysis for system identification and diagnostics (English)
0 references
1991
0 references
The reasoning is based on the heuristic argument that a very small set of observations obtained from a natural dynamic system contains sufficient information about the underlying mathematical dynamic system, even though the observations are of finite precision and contain noise. The authors use time series, filter them for noise in order to `ascertain' the presence of self-generated oscillations, and then estimate some numerical characteristics of the corresponding attractor (mainly its dimension and Lyapunov exponent). Two types of statistical analyses are compared with respect to additive and non-additive noise.
0 references
estimation of attractor dimension
0 references
phase space construction
0 references
Takens theorem
0 references
inhomogeneous attractors
0 references
iterative filtering algorithm
0 references
prediction of time series
0 references
model fitting
0 references
chaotic motion
0 references
dynamical system diagnostics
0 references
noise
0 references
self-generated oscillations
0 references
Lyapunov exponent
0 references
0 references