In: Ajmone Marsan, M.: Lecture Notes in Computer Science, Vol. 691; Application and Theory of Petri Nets 1993, Proceedings 14th International Conference, Chicago, Illinois, USA, pages 166-185. Springer-Verlag, 1993.
Abstract: Deterministic and stochastic Petri nets (DSPNs) are recognized as a useful modeling technique because of their capability to represent constant delays which appear in many practical systems. If at most one deterministic transition is allowed to be enabled in each marking, the state probabilities of a DSPN can be obtained analytically rather than by simulation. We show that the continuous time stochastic process underlying the DSPN with this condition is a Markov regenerative process and develops a method for computing the transient (time dependent) behavior. We also provide a steadystate solution method using Markov regenerative process theory and show that it is consistent with the method of Ajmone Marsan and Chiola.