Red de Detección y Diagnóstico de Fallos de Sistemas Complejos
Escuela2010es.RecentChanges
Edit Page
Page Revisions

Escuela 2010

English version

Noticias

Introducción

Objetivos

Programa

Tasas de Inscripción

Localización



Edita menu

ESCUELA DE DETECCIÓN Y DIAGNOSIS DE FALLOS EN SISTEMAS COMPLEJOS


PROGRAMA PROVISIONAL

T0. CONFERENCIA INAUGURAL (2 horas)


T1. CONCEPTOS FUNDAMENTALES (2 horas)

T1.1. Conceptos: fallos, detección, diagnóstico, fiabilidad,...

T1.2. Principios básicos de los métodos de detección y diagnóstico de fallos, FDI y DX, detectabilidad, observabilidad, diagnosticabilidad,...

T2. DIAGNOSIS BASADA EN MODELOS: PARADIGMA FDI (5 horas)

T2.1. Conceptos de análisis estructural y redundancia analítica.

T2.2. Métodos de detección de fallos basados en modelos: estimación de parámetros, ecuaciones de paridad, observadores de estado aplicado a sistemas lineales y no lineales.

T2.3. Detección de fallos: evaluación de residuos mediante tests de consistencia y generadores de envolventes.

T2.4. Localización de Fallos: residuos estructurados y dirigidos.

T3. DIAGNOSIS BASADA EN MODELOS: PARADIGMA DX (5 horas)

T3.1. Diagnosis basada en modelos, la perspectiva de la Inteligencia Artificial. Diagnosis basada en consistencia (DBC), la propuesta de Reiter.

T3.2. DBC, el GDE y otras aproximaciones computacionales.

T3.3. Diagnosis basada en modelos con restricciones.

T4. BRIDGE: INTEGRACIÓN DE PARADIGMAS FDI y DX (4 horas)

T4.1. Conexiones teóricas y análisis comparativo.

T4.2. Comparación práctica y posibles sinergias.

T5. DIAGNOSIS BASADA EN DATOS (MINERÍA DE DATOS) (3 horas)

T5.1 Diagnosis con métodos estadísticos.

T6. DIAGNOSABILITY ANALYSIS AND RESIDUAL GENERATION (7 horas)

T6.1. Use of structural analysis in the design and analysis of diagnosis systems. Basic structural
analysis algorithm.

T6.2. Residual generation. Outline of approaches for linear and non-linear differential-algebraic models. Modelling for diagnosis in Modelica and Simulink, short demonstration of computer tools for analyzing Modelica models.

T6.3. Placement of sensors to meet certain specified isolability requirements. Industrial application examples of diagnosis techniques.

T7. MODELING AND ALGORITHMS FOR EMBEDDED DIAGNOSTICS (7 horas)

T7.1. Approaches to modeling complex diagnostic systems. Component-based modeling. Modeling discrete, deterministic systems. Modeling discrete, stochastic systems. Modeling discrete/continuous systems.

T7.2. Algorithms for diagnostics inference. "Traditional" MBD algorithms: LTMS, GDE, ATMS. Local-search (approximation) algorithms.

T7.3. Embedded-systems and diagnostics.

Planificación por sesiones:

Comienzo de las clases el lunes 5 de julio y finalización el viernes 9 de julio de 2010. Las sesiones tendrán los siguientes horarios: mañanas de 9.00 h a 13.30 h y tardes de 15.00 h a 18.00 h.

Programa provisional

Time

Monday 5

Tuesday 6

Wednesday 7

Thursday 8

Friday 9

9.00-10.00

T0

T2

T3

T6

T5

10.00-11.00

T0

T2

T3

T6

T5

11.30-12.30

T1

T6

T6

T6

T5

12.30-13.30

T1

T6

T6

T4

T4

Lunch

Lunch

Lunch

Lunch

Lunch

Lunch

15.00-16.00

T2

T3

T7

T4

T4

16.00-17.00

T2

T3

T7

T7

T7

17.00-18.00

T2

T3

T7

T7

T7


INVITED LECTURERS

  • Erik Frisk. Linköpings Universitet, Linköping, Sweden. Ass. Prof. Frisk works at the Department of Electrical Engineering with theory development and application of model based diagnosis techniques in industrial systems. Currently, main research topics include diagnosability analysis of diagnostic models, structural analysis for design of residual generation, observer design for diagnosis, and automotive applications.
  • Gregory Provan. University College Cork, Cork, Ireland. Currently Prof. Provan is head of the Computer Science Department in University College Cork (UCC). He has been technical manager at Rockwell Science Center, where he led projects focusing on control and diagnostics of complex systems, ranging from the Space Shuttle to navy
    ships. Prof. Provan is an expert in model-based diagnostics, discrete-event control, embedded systems, and open-architecture design. Prof. Provan runs an SFI-funded group, the Cork Complex Systems Lab. This research is focused on designing control and analysis methodologies for complex systems-of-systems.
  • Louise Travé-Massuyès, from LAAS-CNRS (Toulouse, France) will talk about BRIDGE: Integration of FDI and DX approaches (T4). She is currently a Research Director of the Centre National de Recherche Scientifique (CNRS), working at LAAS, Toulouse, France, in which she has led the "Qualitative Diagnosis, Supervision and Control" Group for several years. Her main research interests are in qualitative and model based reasoning and applications to dynamic systems monitoring and diagnosis. She is Chairperson of the IEEE SMC Technical Committee on Qualitative Reasoning; member of the IFAC Safeprocess Technical Committee. She is a Senior Member of the IEEE Computer Society. She was also the leader of the MONET2 task group on BRIDGE.

Also, there will be lecturers from the member research groups of the Spanish Network on Supervision and Diagnosis of Complex Systems (Red de Supervisión y Diagnosis de Sistemas Complejos):



Edit Page - Page Revisions - RecentChanges

Si tienes alguna sugerencia, envíala a: rdiag@lsi.us.es
Page last modified on May 26, 2010, at 05:27 PM