scientific article; zbMATH DE number 3886451

From MaRDI portal
Publication:3219219





The author proves a quite general theorem on the asymptotic behavior of solutions to abstract semilinear evolution equations of the type \(du/dt+Lu(t)+Bu(t)=f(t)+h(t)\) where L is a linear and B a monotone operator in a Hilbert space, and f, h are almost periodic. Under suitable assumptions on the equation, especially the existence of an almost periodic solution \(\omega\) in the case \(h=0\), it is shown that in a weak sense any solution of the equation above tends to \(\omega +\zeta\) as \(t\to +\infty\), where \(\zeta\) is a solution of the linear problem \(d\zeta /dt+L\zeta (t)=0.\) If the monotonicity assumption is strong enough one can even show that \(\zeta =0.\) This general theorem is then applied to a number of interesting concrete situations such as semilinear wave equations with damping terms or coupled systems of such equations. A typical result is the following: Consider the mixed problem \[ u_{tt}-\Delta u+\beta (u_ t)=f(t,x)+h(t,x),\quad u(0,x)=u_ 0(x),\quad u_ t(0,x)=u_ 1(x),\quad u|_{\partial \Omega}=0 \] in \((t,x)\in {\mathbb{R}}^+\times \Omega,\) where \(\Omega \subset {\mathbb{R}}^ n\) is a bounded domain, \(\beta\) a continuous non-decreasing function, and \(f,h\in L^ 1({\mathbb{R}}^+,L^ 2(\Omega))\) almost periodic. Assume an almost periodic weak solution \(\omega\) of the equation with \(h=0\) is known to exist. Then for any data \((u_ 0,v_ 0)\in H^ 1_ 0(\Omega)\times L^ 2(\Omega)\) there exists a unique weak solution u(t), and this solution tends to \((\omega +\zeta)(t)\) weakly in \(H^ 1(\Omega)\) as \(t\to +\infty\), where \(\zeta\) satisfies \(\zeta_{tt}-\Delta \zeta =0\). Moreover u'(t) tends to \((\omega '+\zeta ')(t)\) weakly in \(L^ 2(\Omega)\). No growth restrictions on \(\beta\) are necessary. If \(\beta\) is strongly monotone, then \(\zeta =0\). The results are quite general so that they can be applied to many different types of partial differential equations and allow a unified treatment. The problem of asymptotic behavior in the strong sense, i.e. strong convergence, is not attacked. This however is not surprising because even existence of strong solutions of the equations under consideration is essentially unknown.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3219219)