Nonlinear reduction of high-dimensional dynamical systems via neural networks
Authors:
- 21 March 1994
- journal article
- research article
- Published by American Physical Society (APS) in Physical Review Letters
- Vol. 72 (12) , 1822-1825
Abstract
A technique for empirically determining optimal coordinates for modeling a dynamical system is presented. The methodology may be viewed as a nonlinear extension of the Karhunen-Loève procedure and is implemented via an autoassociative neural network. Given a high-dimensional system of differential equations which model a dynamical system asymptotically residing on a stable attractor, the task of the network is to compute a reembedding of the attractor and the dynamics into an ambient space which reflects the intrinsic dimensionality of the problem. The method is demonstrated on the unforced Van der Pol oscillator, the forced Van der Pol, and the Kuramoto-Sivashinsky equation.Keywords
This publication has 6 references indexed in Scilit:
- Iterated Approximate Inertial Manifolds for Navier-Stokes Equations in 2-DJournal of Mathematical Analysis and Applications, 1993
- Some Numerical Aspects of Computing Inertial ManifoldsSIAM Journal on Scientific Computing, 1993
- Reconstructing phase space from PDE simulationsZeitschrift für angewandte Mathematik und Physik, 1992
- Nonlinear principal component analysis using autoassociative neural networksAIChE Journal, 1991
- Approximation by superpositions of a sigmoidal functionMathematics of Control, Signals, and Systems, 1989
- Turbulence and the dynamics of coherent structures. I. Coherent structuresQuarterly of Applied Mathematics, 1987