DEPARTMENTAL SEMINAR

 

 

SPEAKER:

Aswin N. Venkat

 

 
TITLE:

Systems-based Approaches for Optimizing Enterprise-wide Reliability and Economics

 
DATE:

February 24 2011

 
TIME:

10:30 am

 
PLACE:

JHE 326H

 

ABSTRACT

Control of large, networked systems is invariably performed by partitioning the overall system into smaller, more manageable subsystems and applying local modeling and control techniques to each subsystem. Because there is little coordination between the local controllers, they can interact in unex­pected ways not considered in the design phase. As a result, the full system may display fragility and even instability in the face of unmodeled disturbances. An excellent illustration of this phenomenon was the failure of the North American power system in August 2003.
Over the last decade, model predictive control (MPC) has established itself as a premier advanced process control technology. The forecasts produced by MPC provide rich information about future behavior of each subsystem. In this talk, a framework for distributed control of networked systems through the suitable integration of the various subsystems’ MPCs will be described. This talk will fo­cus on strategies that are implementable in practical settings, by taking advantage of the currently de­ployed subsystem models and controllers, and avoiding the exacting modeling requirements and insti­tutional obstacles associated with a centralized control methodology. The key to integrating subsystem-based MPCs is communication of information between subsystems, and cooperation between their con­trollers. The proposed algorithm for integrating subsystem-based MPCs is shown to be more robust than existing methods, with performance approaching that of a fully centralized control framework.
The broader impact of this research lies in the opportunity to demonstrate methods that increase the reliability and economic performance of critical infrastructures that are composed of many highly interacting subsystems. Infrastructure of this type is already becoming ubiquitous, and the need for improving the reliability and economic performance of this infrastructure is increasingly urgent.