On some parabolic boundary control problems with constraints on the control and functional-constraints on the state

From MaRDI portal





The problem of minimizing \(\| w(u;T,\cdot)-z\|_{\infty}\) over all \(u\in L^{\infty}(\Sigma)\) subject to the control constraint \(| u(t,\xi)| \leq 1\) a.e. on \(\Sigma:=]0,T[\times \partial \Omega\) and the state constraints \(c^ 1_ i(t)\leq<\phi_ i,w(u;t,\cdot)>\leq c^ 2_ i(t)\) for all \(t\in [0,T]\), \(i=1,...,k\), is considered. Herein, w(u) is the solution of a parabolic equation with boundary condition \(\beta^{-1}(\partial w(u)/\partial n_ A)+w(u)=u.\) It is shown that under certain assumptions for any optimal control \(u_ 0\) the set \(\{(t,x)\in \Sigma:\quad | u_ 0(t,x)|<1,\quad c^ 1_ i(t)<<\phi_ i,w(u_ 0;t,\cdot)><c^ 2_ i(t),\quad i=1,...,k\}\) has measure zero (generalized bang-bang principle). Among these assumptions a growth condition plays the main rule. This growth condition is analyzed in detail for several special cases.











This page was built for publication: On some parabolic boundary control problems with constraints on the control and functional-constraints on the state

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q788264)