This paper considers the problem of estimation and prediction of chaotic states from arbitrarily nonlinear time series. The basic idea is to use a modified particle filter algorithm to deal with the colored or non-Gaussian noise in chaotic states, the unknown input in chaotic maps, and the nonlinearity in time series. Numerical simulations of Holmes map verify our main results.
ASJC Scopus subject areas
- Statistical and Nonlinear Physics
- Physics and Astronomy(all)
- Applied Mathematics