``Missing moment'' and perturbative methods for polynomial iterated function systems (Q1179058)

The notion of an iterated function system (IFS) was introduced by Barnsley and Demko in 1985, and has foundations in earlier independent work by Hutchinson. An IFS on a compact metric space \(X\) is defined by a finite set of contractive maps \(w_ i: X\times X\) and associated probabilities \(p_ i\), with \(\sum p_ i=1\). The parallel action of the maps \(w_ i\) defines a compact invariant attracting set \(A\subset X\) which supports an invariant measure \(\mu\) and is balanced with respect to the \(p_ i\). When the maps \(w_ i\) on \(X\subset \mathbb{R}\) are linear, the moments \(g_ n=\int x^ nd\mu\) can be recursively computed from the initial value \(g_ 0=1\). However, when the maps \(w_ i\) are nonlinear, the moment relations are incomplete and do not allow recursive computation. The aim of this paper is to describe two methods of accurately estimating the moments when the IFS maps are polynomials. The first method uses Hausdorff conditions to produce bounds on the missing moments. The second method applies perturbation theory to estimate the missing moments from the computed moments of the linearized maps \(w^{(0)}_ i\). The authors apply these two techniques to estimate the missing moments for sample problems.