The model-based fault diagnosability analysis is concerned with the timely detection and isolation of faults by using the system model and observations of the system output. In this paper, we propose the (δ<inf>d</inf>; δ<inf>m</inf>; α)-diagnosability notion for hybrid systems with probabilistic reset, where the faults are diagnosed by observing the timed event sequences. We also present an approach for the analysis of such diagnosability. The (δ<inf>d</inf>; δ<inf>m</inf>; α)-diagnosability notion characterizes the worst- case probability α of detecting and isolating faults within the maximum delay δ<inf>d</inf> since their first occurrence, given the measurement uncertainty δ<inf>m</inf> in observing the time intervals between observed events. We present a method of system abstraction, and prove a quantitative relation between the (δ<inf>d</inf>; δ<inf>m</inf>; α)-diagnosability of the original system and the abstraction. The abstraction has only finitely many trajectories that extend to the end of the time horizon of interest, which allows us to practically calculate the diagnosability and construct the diagnoser.
|Titolo:||Probabilistic diagnosability of hybrid systems|
|Autori interni:||D'INNOCENZO, ALESSANDRO|
|Data di pubblicazione:||2015|
|Appare nelle tipologie:||4.1 Contributo in Atti di convegno|