Convergence of Viable Solutions of Differential Inclusions with Convex Compact Graphs

From MaRDI portal





Let \(z\to G(z)\) be a multivalued mapping in \({\mathbb{R}}^ n\) defined on a bounded domain of \({\mathbb{R}}^ n\) and having a compact and convex graph. A Carathéodory type solution of a differential inclusion (1) dz(t)/dt\(\in G(z(t))\) is said to be viable if it is defined for all positive times. A viable solution z is called an elliptic solution of (1) if it has the form \(z(t)=a \cos \alpha t+b \sin \alpha t\) where \(\alpha\) \(\neq 0\); \(a,b\in {\mathbb{R}}^ n\), \(| a| +| b| \neq 0\). The point \(z\in {\mathbb{R}}^ n\) is a stationary point of G if \(0\in G(z).\) The main result: All the viable solutions of (1) converge, in a uniform rate, to zero, i.e. for every \(\epsilon >0\) there exists \(T>0\) such that \(| z(t)| <\epsilon\) for every \(t>T\) and every viable solution z, iff: (i) zero is the only stationary point of G and (ii) there are no elliptic solutions of (1). It is shown also that the convergence property is generic for the set of differential inclusions with a unique stationary point.











This page was built for publication: Convergence of Viable Solutions of Differential Inclusions with Convex Compact Graphs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3694147)