ABSTRACT
This chapter discusses the system model which includes all the possible faults that are represented by e-transitions (internal transitions that cannot be detected from the discrete event system output). The finite automata (FA) model is transformed into another FA without e-transitions called a diagnoser, in which a test of the system diagnosability is performed. The chapter also discusses the diagnosability characterization of manufacturing systems modelled as interpreted Petri nets (IPNs). It shows a methodology for modelling the normal, permanent, and operational behaviour of a discrete event system and discusses the input–output diagnosability definition. Fault diagnosis is the first stage of a fault recovery system; it is devoted to detect and locate faults, together with the fault confinement and recovery stage that are used to preserve the system integrity and minimize risks to humans, improving the reliability of the system.
