A policy iteration algorithm for fixed point problems with nonexpansive operators |
| |
Authors: | Jean-Philippe Chancelier Marouen Messaoud Agnès Sulem |
| |
Institution: | (1) Ceramics, école Nationale des Ponts et Chaussées, Paritech, 6 et 8 avenue Blaise Pascal, 77455 Marne la Vallée Cedex, France;(2) Inria, Mathfi project, Domaine de Voluceau, BP 105, Rocquencourt, 78153 Le Chesnay Cedex, France |
| |
Abstract: | The aim of this paper is to solve the fixed point problems:
where
is a finite set, L is contractive and B is a nonexpansive operator and
where
and
are general control sets, the operators L
w
are contractive and operators B
z
are nonexpansive. For these two problems, we give conditions which imply existence and uniqueness of a solution and provide
a policy iteration algorithm which converges to the solution. The proofs are slightly different for the two problems since
the set of controls is finite for (1) while it is not necessary the case for problem (2). Equation (2) typically arises in
numerical analysis of quasi variational inequalities and variational inequalities associated to impulse or singular stochastic
control. |
| |
Keywords: | Howard algorithm Policy iteration Impulse control Quasi-variational inequalities Fixed point problems Optimal control of Markov Chains Nonexpansive operators |
本文献已被 SpringerLink 等数据库收录! |
|