Existence of solutions of polynomial-like iterative equation with discontinuous known functions (Q2045213)

The authors study the existence of solutions of polynomial-like iterative equation \(\lambda_{1}f(x)+\lambda_{2}f^{2}(x)+\dots+\lambda_{n}f^{n}(x)=F(x)\), \(x\in I\), with discontinuous known functions. Let \(I=[a, b]\). Then the authors consider the set \[ \mathcal{F}=\{f \mid f : I \rightarrow I \text{ is strictly increasing,} \; f(a)=a \text{ and } f(b)=b\}, \] which contains continuous and discontinuous functions. They define the following three classes of ``good'' functions. For a given \(t\in \operatorname{int} (I)\) define: \begin{gather*} \mathcal{C}_{t}(I) :=\{f \mid f \in\mathcal{F}(I) \text{ is continuous on } I \text{ and } f(t)=t\}, \\ \tilde{\mathcal{C}}_{t}(I) :=\{f \mid f \in\mathcal{F}(I) \text{ is continuous on } \operatorname{int} I \text{ and } f(t)=t\}, \\ \mathcal{H}_{t}(I) :=\{f \mid f \in \mathcal{F}(I) \text{ is discontinuous exactly at point } t \text{ and } f(t)=t\}. \end{gather*} The authors show that if \(f_{1}, f_{2} \in \mathcal{C}_{t}\), then \(f_{1} \circ f_{2} \in\mathcal{C}_{t}\). Similarly, if \(f_{1}, f_{2} \in \mathcal{H}_{t}\), then \(f_{1} \circ f_{2} \in\mathcal{H}_{t}\). Further, the authors introduce piecewise bi-Lipschitz functions and construct a functional space consisting of such functions. They denote the subclasses of all functions \(f\) from \(\mathcal{C}_{t}(I)\) and from \(\mathcal{H}_{t}(I)\) by \(\mathcal{A}_{t}(I, m, M)\) and \(\mathcal{B}_{t}(I, m, M)\) respectively. Moreover, \(\mathcal{G}_{t}(I, m, M)=\mathcal{A}_{t}(I, m, M) \cup\mathcal{B}_{t}(I, m, M)\). They show that the set \(\mathcal{G}_{t}(I, m, M)\) endowed with the distance \(\mathcal{D}(f_{1}, f_{2})=\sup\{|f_{1}(x)-f_{2}(x)|, x\in I\}\) is a complete metric space. Then they define the operator \(\mathcal{T}: \mathcal{G}_{t}(I, m, M) \rightarrow\mathcal{F}(I)\) as \[\mathcal{T}f=\frac{1}{\lambda_{1}}\bigg(F-\sum_{i=2}^{n}\lambda_{i}f^{i}\bigg)\] and prove the existence of solutions by means of the Banach fixed point principle. An example to justify their main results is discussed.