Average cost Markov control processes with weighted norms: Value iteration. (English) Zbl 0829.93068

Gordienko, E.; Hernández-Lerma, O.

Appl. Math. 23, No. 2, 219-237 (1995).

This paper continues the research of the authors [ibid. 23, No.2, 199-218 (1995; Zbl 0829.93070)]. With similar methods and under additional recurrence conditions the convergence of the value iteration algorithm for average cost Markov control processes is proved. It is proved that this result implies average cost optimality and the existence of forecast horizons.

Reviewer: M.Kohlmann (Bonn)

MSC:

93E20	Optimal stochastic control
90C40	Markov and semi-Markov decision processes

Keywords:

Markov control processes; convergence; value iteration; average cost optimality

Citations:

Zbl 0829.93070