Minimax Control of Parabolic Systems with State Constraints

From MaRDI portal





The paper considers the problem of minimizing the functional \[ J(y,u)=\|F(y(\cdot),u(\cdot)\|_{L_\infty(Q)} \] subject to \[ \begin{aligned} & y_t= \text{div }A(x)\nabla u+ f(x,t,y,u)\quad\text{in }Q,\\ & y|_{t=0}= \varphi,\quad y=\phi\quad\text{on }\partial\Omega\times (0,T),\\ & u\in U_{\text{ad}},\quad g(x,t,y(x,t))\leq 0\quad\text{in }\overline Q.\end{aligned} \] Here \(Q= \Omega\times (0,T)\) and \(\Omega\subset \mathbb{R}^n\) is a bounded domain with a smooth boundary \(\partial\Omega\). Under some assumptions (the most essential are: the control set \(U_{\text{ad}}\) is a bounded closed convex subset of \(L_\infty(Q)\); \(f_y'\leq 0\); \(g\) and \(g_y'\) are continuous on \(\overline Q\times\mathbb{R}\)), the authors derive necessary optimality conditions in a Lagrangian form. If \(g\) and \(f\) are linear in \((y,u)\) and \(F\) is convex in \((y,u)\) then these conditions are necessary and sufficient. To overcome the difficulty that the adjoint equation can involve finitely additive measures, the authors give a special representation theorem for such measures from \((L_\infty(Q))^*\).











This page was built for publication: Minimax Control of Parabolic Systems with State Constraints

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4943722)