Robust fault detection of singular LPV systems with multiple time-varying delays

In this paper, the robust fault detection problem for LPV singular delayed systems in the presence of disturbances and actuator faults is considered. For both disturbance decoupling and actuator fault detection, an unknown input observer (UIO) is proposed. The aim is to compute a residual signal whi...

Descripción completa

Detalles Bibliográficos
Autores: Hassanabadi, Amir Hossein, Shafiee, Masoud, Puig Cayuela, Vicenç|||0000-0002-6364-6429
Tipo de recurso: artículo
Fecha de publicación:2016
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/87177
Acceso en línea:https://hdl.handle.net/2117/87177
https://dx.doi.org/10.1515/amcs-2016-0004
Access Level:acceso abierto
Palabra clave:Nonlinear systems
System theory
singular delayed LPV systems
fault detection
unknown input observer (UIO)
robustness
fault sensitivity
unknown input observers
detection filter design
descriptor systems
stability
stabilization
generation
Sistemes no lineals
Sistemes, Teoria de
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Descripción
Sumario:In this paper, the robust fault detection problem for LPV singular delayed systems in the presence of disturbances and actuator faults is considered. For both disturbance decoupling and actuator fault detection, an unknown input observer (UIO) is proposed. The aim is to compute a residual signal which has minimum sensitivity to disturbances while having maximum sensitivity to faults. Robustness to unknown inputs is formulated in the sense of the H-infinity-norm by means of the bounded real lemma (BRL) for LPV delayed systems. In order to formulate fault sensitivity conditions, a reference model which characterizes the ideal residual behavior in a faulty situation is considered. The residual error with respect to this reference model is computed. Then, the maximization of the residual fault effect is converted to minimization of its effect on the residual error and is addressed by using the BRL. The compromise between the unknown input effect and the fault effect on the residual is translated into a multi-objective optimization problem with some LMI constraints. In order to show the efficiency and applicability of the proposed method, a part of the Barcelona sewer system is considered.