Fault diagnosis method for timed discrete-event systems: Application to autonomous electric vehicle

International audienceWith the explosion of new technologies, a new type of systems has been designed. Their characteristics make them more and more complex, because of the increasing number of components and the mix of technologies (electronic, mechanic, software). In the field of functional safety, this creates problems to apply the existing methods. In particular, the fault diagnosis can be long and difficult due to these methods, which are not adapted anymore. In this paper, a fault diagnosis method is proposed for timed discrete- event systems, based on a timed automaton as a model of the system. Two concepts are used: the failure detection using only some discrete characteristic times and the isolation with the fault signatures. At first every stage of the modeling are explained, through the example of an autonomous electric vehicle. Then the diagnosis method is described step by step on this example. The efficiency is finally discussed