Application of Timed Automata to the Diagnosis of the DAMADICS Benchmark Problem

The paper describes a design method of a diagnostic algorithm for detecting and identifying faults that occur in the discrete–event systems described by timed automata. The timed automata used for diagnosis do not only incorporate only the corresponding states and transitions into normal operation behaviour of the systems but also additional states and transitions representing the faulty behaviour of them. The use of timed automata improves diagnostic performance by providing temporal information on how the systems behave. The results of the diagnosis of timed automata consist of two parts. The first part detects the existence of a fault in the system while the second is the fault identification. The method is applied to detect the faults in the industrial actuator in the Lublin sugar factory. Copyright c ©