You can only gain access to certain items and special pricing if you have logged in. Login Now.

CC-03-164 How Much Can We Increase the Efficiency of MPC Identification?

Yucai Zhu President Tai-Ji Control The Netherlands Jay Deshmukh President JD Control Systems Sugar Land, TX Ralph Lien Senior Process Control Engineer Rohm and Haas Company Deer Park, TX Location

Format:
Electronic (digital download/no shipping)

Associate Member, International Member, Petrochemical Member, Refining Member - $0.00
Government, NonMember - $25.00

Description:

In refining and chemical industries, there is a growing demand for efficient model identification methods in MPC commissioning and maintenance. In this work we will introduce a modern identification technology called “Tai-Ji ID”. Key components of the method are: 1) automatic and multivariable tests, can be open loop or closed-loop, 2) automatic parametric model estimation and 3) automatic model validation/selection. The advantages of the technology include reduced test time (70%), reduced process disturbance, accurate models and user-friendliness. An application of the technology at Rohm and Haas is reported. The Tai-Ji ID was chosen to do the re-identification for a multivariable controller that has been successfully in operation for over a year. Initially, the existing models were identified using traditional step test data. The step test lasted for many days around the clock. The identification test with Tai-Ji ID for the dynamic models took a shorter duration where all the eight MVs were moved (stepped) simultaneously. Part of the inferential models was also identified using the multivariable test data. The reduction of test time by over 70% was realized compared to the initial installation. The tests had not effected unit operation. The performance of the controller with the new model is satisfactory.

Product Details:

Product ID: CC-03-164
Publication Year: 2003