Estimation methods for passage times using one-dependent cycles (Q1911470)

This paper deals with simulation methods that are applicable to delays when the system under study is modeled as a discrete-event stochastic system. Simulation is often the only available means for analyzing the length of a sequence of delays. It is sometimes possible to obtain meaningful estimates for the limiting average delay indirectly, without measuring lengths of individual passage time intervals. The authors consider sequences of delays determined by state transitions of a generalized semi-Markov process (G.S.M.P.) and introduce a recursively-generated sequence of real valued random vectors, to provide the relation between the starts and determinations of passage time intervals. They show, for example, that if the G.S.M.P. has a recurrent state space, then the sample paths of any sequence of delays can be decomposed into dependent identically distributed cycles. Applications of these methods to a manufacturing unit with robots is discussed.