scientific article; zbMATH DE number 1099130

From MaRDI portal
Publication:4369318





Let \(\Omega\) be an open bounded subset of \(\mathbb{R}^N\) \((N\geq 2)\) with a boundary \(\Gamma\) of class \(C^\infty\), let \(Q= \Omega \times(0,T)\) and \(\Sigma= \Gamma \times (0,T)\) for some \(T>0\). Then the following hyperbolic initial boundary value problem is considered in this paper: \[ {\partial^2 y\over \partial t^2} +Ay= a(x,t,u) \text{ in }Q, \tag{1} \] \[ {\partial y\over \partial n_A} (s,t)= b(s,t,v) \text{ on }\Sigma, \tag{2} \] \[ y(x,0)= y_0(x), \quad {\partial y\over \partial t} (x,0)= y_1(x),\;x\in \Omega, \tag{3} \] with \(A\) being a uniformly second order elliptic operator with regular coefficients only depending on \(x\in\Omega\), \({\partial \over \partial n_A}\) being the normal derivative relative to \(A\), with \(y_0\in H^2 (\Omega)\), \({\partial y_0 \over \partial n_A} =0\), and \(y_1\in H^1 (\Omega)\). The functions \(u\) and \(v\) are control functions taken from the set \[ U_{ad} =\bigl\{u\in L^p(Q) \mid u(x,t)\in K_1(x,t) \text{ a.e. in } Q\bigr\} \] and \[ V_{ad} =\bigl\{v\in L^p (\Sigma) \mid v(s,t)\in K_2(x,t) \text{ a.e. in }\Sigma \bigr\}, \] respectively. The problem to be solved consists of finding \(u\in U_{ad}\) and \(v\in \Sigma_{ad}\) such that the cost functional \[ \begin{aligned} J(y,u,v)= & \int_Qf_1 \bigl(x,t,y(x,t) \bigr)dx dt +\int_Q g_1\bigl(x,t,u (x,t)\bigr) dx dt\\ & +\int_\Sigma f_2\bigl(x,t,y (s,t)\bigr) ds dt+ \int_\Sigma g_2 \bigl(s,t,v(s,t) \bigr)ds dt \end{aligned} \] is minimized where \(y=y(x,t)\) is the corresponding weak solution of (1), (2), (3). Under suitable assumptions on \(a,b,K_1, K_2, f_1,f_2,g_1\) and \(g_2\) it is shown that this problem has a solution, if \(f_1(x,t, \cdot)\) is concave for almost all \((x,t)\in Q\) and \(f_2(s,t, \cdot)\) is concave for almost all \((s,t) \in \Sigma\). The idea of the proof is to associate with this problem a convexified problem and to use a variant of the Lyapunov convexity theorem. In Section 3 of the paper related existence theorems are considered and in Section 4 a bang-bang-principle for problems governed by the one dimensional wave equation is given together with the main idea of the proof. For the entire collection see [Zbl 0877.00012].











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 Q4369318)