No label defined (Q4786560)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1846819
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 1846819

      Statements

      0 references
      0 references
      17 December 2002
      0 references
      exit time
      0 references
      optimal control
      0 references
      value function
      0 references
      semiconcavity
      0 references
      Hamilton-Jacobi equation
      0 references
      viscosity solution
      0 references
      maximum principle
      0 references
      Continuing the previous research in \textit{P. Cannarsa}, \textit{C. Pignotti} and \textit{C. Simnstrari} [Discrete Contin. Dyn. Syst. 6, No. 4, 975-997 (2000)] the authors prove a large number of interesting results concerning semiconcavity of the value function, \(v(.)\), of the ``exit-time'' problem which consists in the minimization of the functionals: \[ J(x,\alpha(.)):=\int_0^\tau e^{-\lambda t}L(y(t)) dt, \;x\in \Omega \subset \mathbb{R}^n, \] subject to: \[ y'(t)=\alpha(t)\in B:=\{a\in \mathbb{R}^n; \;|a|\leq 1\} \text{ a.e. }(0,\tau), \;y(0)=x, \] where \(\tau=\tau (x,\alpha(.))\) is ``the exit-time'', at which the trajectory \(y(.)=y_x^{\alpha(.)}(.)\) reaches the boundary, \(\partial \Omega\), of the open subset \(\Omega\subset\mathbb{R}^n\). Most of the results in the paper rely on the fact that if \(L(.):\overline \Omega \to \mathbb{R}_+:=[0,\infty)\) is Lipschitz and \(L(x)= 0 \;\forall \;x\in \partial\Omega\), then the value function of the problem may be represented as follows: \[ v(x)=\inf\{\int_0^\infty e^{-\lambda t}L(y(t)) dt; \;y'(t)\in B \text{ a.e., } y(0)=x, \;y(t)\in \overline \Omega \;\forall \;t\geq 0\}, \;x\in \overline \Omega:=\Omega\cup \partial \Omega, \] and also on certain techniques of construction of admissible trajectories to the second problem. The main result of the paper is Theorem 4.1 (whose very technical proof takes some 16 pages) stating, essentially, that under some additional (rather restrictive) hypotheses, the value function is semiconcave. Moreover, in Section 3 it is proved (under milder hypotheses) that the value function is Hölder continuous while in Section 5, as applications of the semiconcavity, several other results are proved concerning differentiability along optimal trajectories, the Maximum Principle and its relations with the value function, characterizations of the optimal trajectories and propagation of singularities of the value function.
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references
      0 references