About

Model Predictive Control (MPC) traditionally finds use on a supervisory level of process control systems. Different vendors offer layered packages to interface with existing control systems. These solutions often require rather large commissioning and configuration efforts introduced by the overhead for interface and communication. This paper provides a summary of benefits and challenges of optimizing and embedding an MPC algorithm as a Function Block (FB) at the controller level. Representing this very sophisticated control strategy with the FB application model results in ease of use and consistency with database, redundancy, alarm and user interface management of a control system. The effective use of field mounted controllers to host the control application allows control of processes with time constants in the range of seconds. Leveraging off of the mode and status paradigm of function blocks, safe data transfer can be insured. Automatic data acquisition and process model generation is discussed. Using function blocks that can be wired in graphical editors it becomes very easy to replace existing multivariable control strategies based on PID networks.

- Downloadable files require Adobe Acrobat Reader.

Customers who bought this product also bought: A Simplified and Integrated Approach to Model Predictive Control Implementation - ISA EXPO 2000 Simulation of Multivariable Model Predictive Control - ISA 2001 Stable Nonlinear Model Predictive Control of Constrained Wiener and Hammerstein Models - INTERKAMA 1999 Advanced Control Unleashed: Plant Performance Management for Optimum Benefit Control System Documentation: Applying Symbols and Identification, 2nd Edition