Programma Optimal Control + Lab

Onderwerpen

Identification and modeling: (Perform a complete identification of a high purity distillation column)

• Freerun, Step test, PRBNS test
• PRBNS test design 
• Data Pre-Processing (filtering, peak shaving, resampling, detrending) 
• Model Identification (FIR) 
• Model validation

Model Predictive Control: (Design and implementation of a controller for a high purity distillation column) 

• MPC basics: prediction, static and dynamic optimization, receding horizon
• Zones and ideals (setpoints) 
• Control hierarchy and priorities 
• Tuning parameters of a MPC controller

IPCOS Integrated Software Platform (ISP): flexible, open, software architecture, which allows the user to implement a Model Predictive Controller with all needed functionality for industrial processes. 

• OPC communication (OPC for Matlab)

Competenties

• Understanding what kind of control problem is ideally suited for Model Predictive Control.
• Understanding the different steps needed for successful implementation of a Model Predictive Controller.
• Understanding the MPC basics. 
• Be familiar with the main tuning parameters of a Model Predictive Controller. 
• Become familiar with one of the available commercial MPC software packages on the market and see how theory is used in practice.

Doelstellingen

• The different steps needed to successfully implement a Model Predictive Controller.
• Multivariable dynamic process behavior and its impact on process control.
• Model Predictive Control and the design of such a controller.
• Obtain the skills to use a commercial available Model Predictive Control software package for solving an optimal control problem on an industrial process unit.