Global reconstruction from nonstationary data

One-dimensional time series of a dynamical system with slowly varying parameters are investigated. For estimation of the characteristics of the attractors of such a system which exist for fixed values of the parameters, it is proposed to “cut out” from the time series short segments that belong to the individual attractors and to use them to reconstruct a model dynamical system. Pis’ma Zh. Tekh. Fiz. 25, 74–81 (May 26, 1999)