Exploiting T-invariant analysis in Diagnostic reasoning on a Petri net model

This paper copes with the application of T-invariant analysis to diagnostic reasoning based on a Petri net model. In particular, it is formally shown how the notion of diagnostic solution can be related to that of Petri net T-invariant, partially transforming a problem traditionally solved by means of symbolic techniques into a linear algebraic one. The approach has been inspired by some previous works concerning the use of T-invariant analysis on Petri net models of logic programs and it takes its place among recent approaches aiming at integrating artificial intelligence and Petri netbased techniques. A diagnostic algorithm exploiting the idea is proposed and the role of the Petri net model in the resulting diagnostic architecture is discussed.