A fault diagnosis approach for hybrid systems

This communication presents a first attempt in the use of a dynamic hybrid simulator within a fault diagnosis system coupled to a classification methodology. The use of hybrid modelling has the advantage to clearly separate the continuous aspects from the discrete ones; this allows an analysis of causalities resulting from the state changes. Our approach begins with the simulation of the system in normal conditions until the existence of dysfunction has been detected (by comparing process measurement and simulation result). Then, a backward simulation through the chain of causality is performed using the same Petri net and the method is propagated until the coherence between simulation and real behaviour is reached. Nevertheless, according to this procedure all the possible scenarios of faults should be explored. A classification method based on measurement data would be used to restrict the tree of possibilities of faults to be explored. In the framework of this study, the simulation aspects have been entrusted to the general object-oriented environment PrODHyS (process object dynamic hybrid simulator). Its major characteristic is its ability to simulate systems described with object differential Petri nets (ODPN) formalism