Authors:
Paolino Tona
1
and
Jean-Marc Bader
2
Affiliations:
1
Institut Français du Pétrole, France
;
2
Axens, France
Keyword(s):
System identification, model predictive control.
Related
Ontology
Subjects/Areas/Topics:
Informatics in Control, Automation and Robotics
;
Signal Processing, Sensors, Systems Modeling and Control
;
System Identification
Abstract:
ISIAC (as Industrial System Identification for Advanced Control) is a new software package geared to meet the requirements of system identification for model predictive control and the needs of practicing advanced process control (APC) engineers. It has been designed to naturally lead the user through the different steps of system identification, from experiment planning to ready-to-use models. Each phase can be performed with minimal user intervention and maximum speed, yet the user has every freedom to experiment with the many options available. The underlying estimation approaches, based on high-order ARX estimation followed by model reduction, and on subspace methods, have been selected for their capacity to treat the large dimensional problems commonly found in system identification for process control, and to produce fast and robust results. Models describing parts of a larger system can be combined into a composite model describing the whole system. This gives the user the fle
xibility to handle complex model predictive control configurations, such as schemes involving intermediate process variables.
(More)