Adaptive Fault Diagnoser based on PSO algorithm for a class of Timed Continuous Petri Nets

This work is concerned with the implementation of an Adaptive Fault Diagnoser (AFD) for a system modeled by Timed Continuous Petri Nets under infinite server semantics, where the set of potential faults is a priori known, however their presence during system evolution, type, location, occurrence time, magnitude and behavior over time are unknown. There exist previous works reported in literature, where this problem has been solved, unfortunately the number of diagnosers used to detect, isolate and identify the fault is too large. Now, this work proposes a single diagnoser model where its structure is known and some of its parameters are updated depending on the fault occurrence. Considering this model, identification algorithms, based on heuristic optimization methods, are used to identify these unknown fault parameters. The analysis of the diagnoser parameters allows the faults detection, isolation and identification. The effectiveness of the proposed diagnoser is shown through two examples with different fault behaviors.