Premium
Practical realizations in process modeling: Part I. A new approach to practical realization
Author(s) -
Jayaraman Krishnamurthy,
Lapidus Leon
Publication year - 1976
Publication title -
aiche journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.958
H-Index - 167
eISSN - 1547-5905
pISSN - 0001-1541
DOI - 10.1002/aic.690220212
Subject(s) - realization (probability) , smoothness , process (computing) , markov process , mathematical optimization , markov chain , algorithm , computer science , mathematics , state (computer science) , control theory (sociology) , artificial intelligence , machine learning , statistics , mathematical analysis , control (management) , operating system
Abstract A new approach is presented for the construction of linear, low‐order stable, state‐variable, dynamic models from experimental pulse response data of process systems: the problem of practical realization. This problem is solved by deriving and implementing constraints in the estimation of Markov parameters which are input to a realization algorithm. This is accomplished sequentially by using an algorithm for recursive partial realization. Bounds on the Markov parameters are derived from qualitative information about the smoothness of the response curve and used in a stagewise regression procedure. The presente procedure identifies the faster modes first; hence the tail of the pulse response, if inordinately long and inaccurate, may be ignored while lower‐order partial realizations are constructed.