Takens embedding theorems for forced and stochastic systems (Q4378786)

The main purpose of their paper is to extend Takens's Embedding theorem both to deterministically forced systems and to stochastic systems, which are more present in the researchers' attention than the deterministic systems. The study begins with an overall description of the usual framework for studying time series generated by deterministic dynamical systems and after this the authors give a statement of the standard Takens Theorem and briefly discuss some of its consequences. They draw the conclusion that it is important to develop generalizations of the Takens Theorem applicable to stochastic systems and to forced systems. This is because both Takens's Theorem and the Kupka-Smale Theorem require the dynamics to be deterministic and autonomous. In the second section, the authors assume that the smooth map \(f\) (which describes the dynamics) is forced by some second system \(g\) and they study the forced dynamics of \(f\) by expanding the phase space and by incorporating the forcing dynamics. The authors show that one can typically reconstruct the joint dynamics of an unknown forcing and a forced system from a time series of an observation of the forced system alone, through Theorem 3 (Forced Takens Theorem, Stark) and Theorem 4 (Bundle Belay Embedding Theorem, Stark). The last section, ``Stochastic Systems'' treats this wide range of systems and leads to the Takens Embedding Theorem for Stochastic Systems. The accurate study included in this paper offers a precious instrument for the study of deterministically forced systems and stochastic systems, which appear in most practical applications.