Discrete-event modeling can be applied to a large variety of physical syste
ms, such as digital hardware, quelling networks, communication networks, an
d industrial protection systems, in order to support different tasks, inclu
ding fault detection, monitoring, and diagnosis. This paper focuses on the
model-based diagnosis of a class of distributed discrete-event systems, cal
led active systems. An active system, which is designed to react to possibl
y harmful external events, is modeled as a network of communicating automat
a, where each automaton describes the behavior of a system component. Unlik
e other approaches based on the synchronous composition of automata and on
the off-line creation of the model of the entire system, the proposed diagn
ostic technique deals with asynchronous events and does not need any global
diagnoser to be built. Instead, the current approach features a problem-de
composition/solution-composition nature whose core is the on-line progressi
ve reconstruction of the behavior of the active system, guided by the avail
able observations. This incremental technique makes effective the diagnosis
of large-scale active systems, for which the one-shot generation of the gl
obal model is almost invariably impossible in practice. The diagnostic meth
od encompasses three steps: 1) reconstruction planning; 2) behavior reconst
ruction; and 3) diagnosis generation, Step 1 draws a hierarchical decomposi
tion of the behavior reconstruction problem. Reconstruction is made up in S
tep 2, where an intensional representation of ail the dynamic behaviors whi
ch are consistent with the available system observation is produced. Diagno
sis is eventually generated in Step 3, based on the faulty evolutions incor
porated within the reconstructed behaviors. The modular approach is formall
y defined, with special emphasis on Steps 2 and 3, and applied to the power
transmission network domain.