Building diagnosable Petri net models for distributed fault location of DES

The paper deals with diagnosability and fault location of discrete event systems (DES). It presents a modular method for building diagnosable models using interpreted Petri nets, and proposes an efficient strategy for computing faulty markings. Model construction is based on a modular methodology in which the modules are PN state machines where the diagnosability analysis is polynomially performed. Then several rules for module composition are defined in order to preserve the diagnosability property in the global model