Premium
A learning fedforward control strategy for the concentration control of recycle reactors. Part I: Set‐point control
Author(s) -
Schnitzlein Klaus,
Löwe Arno
Publication year - 1992
Publication title -
chemical engineering and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.403
H-Index - 81
eISSN - 1521-4125
pISSN - 0930-7516
DOI - 10.1002/ceat.270150106
Subject(s) - feed forward , control (management) , task (project management) , set (abstract data type) , point (geometry) , control theory (sociology) , control system , computer science , control engineering , engineering , process engineering , mathematics , artificial intelligence , geometry , systems engineering , electrical engineering , programming language
Unknown reaction rates make concentration control in laboratory recycle reactors a difficult task, particularly when gas chromatographs or other analytical equipment with long analysis times from part of the control system. For such time‐delay systems, a model‐based feedforward control technique is developed which uses some kind of learning linear mapping to provide estimates of the reaction rates. Regarding the number of time steps necessary to satisfy a new set point, in a variety of simulation studies, this technique has proven superior to similar but simpler control strategies.