scientific article; zbMATH DE number 1500621

From MaRDI portal
Publication:4501643





As stated in the Introduction, the aim of this paper is to extend to \textit{infinite horizon optimal control problems} authors' previous results in [\textit{H. Frankowska} and \textit{S. Plaskacz}, ``Semicontinuous solutions of Hamilton-Jacobi equations with state constraints in J. Andres (ed.) et al., Differential inclusions and optimal control, Lecture Notes in Nonlinear Analysis, Vol. 2, J. Schauder Centre for Nonlinear Studies, 145-161 (1998)]. Most of the results in the paper concern the value function \[ V(x_0):=\inf_{u(.)\in {\mathcal A}(x_0)}\int_0^\infty e^{-s}L(x(s),u(s)) ds \] where for each \(x_0\in K\subset R^n, \;{\mathcal A}(x_0)\) denotes the set of all (admissible) measurable controls \(u(.):R_+\to U\) for which the (unique) solution \(x(.)=x(.;x_0,u(.))\) of the differential equation \(x'=f(x,u(t), \;x(0)=x_0\) is defined on \(R_+=[0,\infty)\) and satisfies the state constraints \(x(t)\in K \;\forall \;t\geq 0\). Assuming, among other hypotheses, the fact that \(K=Cl(D)\) where \(D\subset R^n\) is a locally compact subset and replacing Soner's ``controllability condition by a different one expressed in terms of a relative paratingent cone, the authors prove in their main result, Theorem 6, that \(V(.)\) is the unique lower semicontinuous generalized solution (of viscosity type) of the associated Hamilton-Jacobi equation: \[ V(x)+H(x,DV(x))=0, \;x\in K, \;H(x,p):=\sup_{u\in U}[\langle f(x,u),p\rangle-L(x,u)]. \] The rather long and technical proof of the main result is based on the up-to-date versions of the theorems concerning the viability and, respectively, the invariance of solutions of autonomous differential inclusions and also on a large number of auxiliary results. For the entire collection see [Zbl 0921.00020].




Cited in
(19)








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 Q4501643)