State uncertainty analysis of Fuzzy Timed DES

Abstract

This paper addresses state estimation of discrete event systems (DES) exhibiting variations on the duration of activities; the system state is approximated by the marking of Fuzzy Timed Petri Net (FTPN) models. For this purpose a definition of FTPN is presented in which fuzzy sets associated to places represent the ending time uncertainty of activities. A fuzzy marking matrix equation is proposed, allowing computing the marking evolution of cyclic FTPN. Finally, the uncertainty of the estimated marking is evaluated as a measure of quality on the approximation.