Premium
Centralized vs decentralized adaptive generalized predictive control of a biodiesel reactor
Author(s) -
Ho Yong Kuen,
Mjalli Farouq S.,
Yeoh Hak Koon
Publication year - 2012
Publication title -
asia‐pacific journal of chemical engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.348
H-Index - 35
eISSN - 1932-2143
pISSN - 1932-2135
DOI - 10.1002/apj.1638
Subject(s) - model predictive control , controller (irrigation) , control theory (sociology) , recursive least squares filter , decentralised system , adaptive control , process (computing) , control (management) , biodiesel , computer science , servo , scheme (mathematics) , control engineering , engineering , mathematics , artificial intelligence , algorithm , adaptive filter , chemistry , mathematical analysis , biochemistry , agronomy , catalysis , biology , operating system
A second look at biodiesel reactor control using Recursive Least Squares (RLS)‐based adaptive Generalized Predictive Control (GPC) strategy revealed the possibility of a simpler alternative to the previously published centralized RLS‐based GPC controller (CRLS‐GPC). New results show that the simpler decentralized RLS‐based GPC controller (DRLS‐GPC) was on par with the more sophisticated centralized version in terms of servo and regulatory control, process interactions handling, and the resultant controller moves. Moreover, the simplified control scheme remained superior to the conventional Proportional–Integral controller. Such attributes make the DRLS‐GPC an attractive compromise between complexity and performance. © 2012 Curtin University of Technology and John Wiley & Sons, Ltd.