This article is cited in 3 papers

Research articles

Dynamic programming algorithms for solving stochastic discrete control problems

Dmitrii Lozovanu^a, Stefan Pickl<sup>b

a</sup> Institute of Mathematics and Computer Science, Academy of Sciences of Moldova, Chişinău, Moldova
^b Institut für Angewandte Systemwissenschaften und Wirtschaftsinformatik, Fakultät für Informatik, Universität der Bundeswehr, München

Abstract: The stochastic versions of classical discrete optimal control problems are formulated and studied. Approaches for solving the stochastic versions of optimal control problems based on concept of Markov processes and dynamic programming are suggested. Algorithms for solving the problems on stochastic networks using such approaches and time-expended network method are proposed.

Keywords and phrases: time-discrete system, optimal control, stochastic networks, markov processes, dynamic programming.

MSC: 93E20, 49L20

Received: 25.01.2009

Language: English

Bibliographic databases:

• 
•