In this section
@inproceedings{Boem:2016:10.1109/CDC.2016.7798443,
author = {Boem, F and Carli, R and Farina, M and Ferrari-Trecate, G and Parisini, T},
doi = {10.1109/CDC.2016.7798443},
publisher = {IEEE},
title = {Scalable monitoring of interconnected stochastic systems},
url = {http://dx.doi.org/10.1109/CDC.2016.7798443},
year = {2016}
}
TY - CPAPER
AB - In this paper, we propose a novel distributed faultdetection method to monitor the state of a linear system, par-titioned into interconnected subsystems. The approach hingeson the definition of a partition-based distributed Luenbergerestimator, based on the local model of the subsystems andthat takes into account the dynamic coupling terms betweenthe subsystems. The proposed methodology computes –in adistributed way– a bound on the variance of a properly definedresidual signal, considering the uncertainty related to thestate estimates performed by the neighboring subsystems. Thisbound allows the computation of suitable local thresholds withguaranteed maximum false-alarms rate. The implementationof the proposed estimation and fault detection method isscalable, allowingPlug & Playoperations and the possibilityto disconnect the faulty subsystem after fault detection. Theo-retical conditions guaranteeing the convergence of the estimatesand of the bounds are provided. Simulation results show theeffectiveness of the proposed method.
AU - Boem,F
AU - Carli,R
AU - Farina,M
AU - Ferrari-Trecate,G
AU - Parisini,T
DO - 10.1109/CDC.2016.7798443
PB - IEEE
PY - 2016///
TI - Scalable monitoring of interconnected stochastic systems
UR - http://dx.doi.org/10.1109/CDC.2016.7798443
UR - http://hdl.handle.net/10044/1/43605
ER -