Diagnosis of discrete event system by stochastic timed automata

the paper describes a diagnosis method for detecting and identifying faults that occur in a discrete-event system described by a stochastic timed automaton. The proposed diagnosis algorithm is based whether the measured event time sequences are consistent with a stochastic timed discrete event model. The degree of this consistence is given by a continuous diagnosis function. The diagnosis which consists of two parts (fault detection and identification) is then obtained.