Home Up Reach Us Contents                             

ITOBO Research  

Provan's Research ITOBO Research

Home
Up
Members
Research
Publications

ITOBO

This SFI-funded research activity focuses on the analysis of complex systems, as described as systems of systems. In particular, we are interested in defining formal models, and designing efficient inference algorithms for such systems.

Focus Areas

bullet

Modelling Integrated Building Systems (IBS)

bulletIntegrated Systems Analysis and Middleware
bulletInference algorithms for IBS inference
bulletApplication areas: hybrid systems modelling, control of lighting and HVAC

Members

bullet Director: Prof. Gregory Provan
bullet Postdoctoral fellows:  Menouer Boubekeur
bulletPhD students: TBD

Areas of Technical Analysis

Modeling: My work in building systems representation has focused on hybrid systems models for building sub-systems

languages for system simulation, analysis and control. In particular, I have studied structure-based representations (which include Bayesian networks, Influence Diagrams, Causal networks, and constraint satisfaction problems), logical representations, and control representations (Finite State Automata and Petri nets). One area of particular interest is the development of a representation that can perform simulation, diagnosis and control reconfiguration, since such a representation is relevant to a large number of important applications. Recently, my primary focus has been on how to define distributed representations for distributed networked systems.

Systems Integration: Software tools for complex systems require not only efficient algorithms but algorithms that work together in a coordinated fashion, which requires architecture design, middleware and systems analysis. I have focused on open-architecture design, and built open-architecture frameworks for applications like health monitoring and control reconfiguration. On the middleware front, my work has addressed “adaptive, reflective” middleware, which incorporates systems monitoring and resource allocation within the middleware. On the systems analysis front, my work has addressed content-based QoS, probabilistic analysis and resource allocation issues.

Inference: My work in inference has focused on designing efficient algorithms tailored to specific representations and specific application platforms (which typically have limited CPU and memory). Much of my work has been devoted to designing efficient algorithms for probabilistic inference, and for diagnostic inference. I have extended this work to include algorithms for distributed diagnostic inference. In addition, I have designed algorithms that integrate control and diagnosis.

Publications

Publications generated under this project are described here.

 

Send mail to ccsl@cs.ucc.ie with questions or comments about this web site.
Copyright © 2008 Cork Complex Systems Lab
Last modified: 05/26/08