Filippov-Pliss lemma and \(m\)-dissipative differential inclusions (Q2393056): Difference between revisions

The authors consider a differential inclusion in a Banach space \(X\) of the form \[ \dot{x}(t) \in Ax(t) + F(t,x(t)), \quad x(0) = x_0 \in \overline{D(A)}, \tag{1} \] where \(A\) is an \(m\)-dissipative operator and \(F\) is an almost upper hemicontinuous multimap with nonempty convex weakly compact values, satisfying a one-side Peron and a sublinear growth condition. A continuous function \(x(\cdot)\) is called an \(\varepsilon\)-solution of (1) if there exists a function \(g(\cdot)\) with \(\|g(t) - x(t)\| < \varepsilon\) such that \(x(\cdot)\) is a solution of \(\dot{x} \in Ax + F(t,g(t)), \,\,x(0) = x_0.\) The authors present the following version of the Filippov-Pliss lemma. There exists a positive continuous function \(\nu(\cdot)\) with \(\nu(0) = 0\) such that for every \(\varepsilon\)-solution \(x(\cdot)\) of (1) there exists a solution \(y(\cdot)\) of (1) such that \(\|x(t) - y(t)\| \leq \nu(\varepsilon).\) This result is applied to prove a Kneser type result about the connectedness of the solution set and to justify the existence of an attractor for an autonomous inclusion in which the multivalued term satisfies a Lipschitz-type condition with a negative constant.