Diagnostic of discrete event systems using timed automata in MATLAB SIMULINK

Generally, diagnostic involves two interrelated phases: the detection and the localization of failures. The approach proposed in this paper is based on operating time. The method is applicable to any system whose dynamical evolution depends not only on the order of discrete events but also on their durations as in industrial processes. Diagnosis of faults is achieved through the implementation of a model observer based on timed automata. This observer called diagnoser makes it possible to detect and locate possible failures in real time. A failure is detected when an event is not reaching at the waited date, or if it last too long compared to its ongoing operations. Temporal knowledge of the process to be monitored is therefore essential. The pro-posed diagnoser is a monitoring tool that can detect, isolate and locate a fault in a system. The used method-ology is based on the timed automata. The presence of an error corresponds to the execution of a state defined as the defective controller. For the detection phase, parameter detectability is the ability to detect a fault in the system. For the localization phase, the isolation is a property that corresponds to the ability to isolate (locate) a fault. The diagnostic performance is quantified through two parameters: the detection time and isolation time.