A HYBRID MODELING APPROACH TO IDENTIFY A CONTINUOUS PROCESS IN A CLOSED LOOP
作者:
HSIAO-PING HUANG,
YUNG-CHENG CHAO,
期刊:
Chemical Engineering Communications
(Taylor Available online 1983)
卷期:
Volume 22,
issue 5-6
页码: 345-369
ISSN:0098-6445
年代: 1983
DOI:10.1080/00986448308940067
出版商: Taylor & Francis Group
数据来源: Taylor
摘要:
A hybrid modeling approach is proposed for modeling and identification of an uniformly sampled continuous process in a closed loop where noises exist. This approach combines the determination of an ARMAX model with the determination of a continuous model as an integrated modeling procedures. The resulted model is designated as the HM(n,m,s) model. By using this modeling approach, the identifiability of the HM(n,m,s) model will be regardless to the regulator that is being used in the closed loop. Besides, the HM(n,m,s) model provides more accurate dynamic information than those of the conventionally estimated ARMAX model, which may be quite misleading sometimes. A systematic modeling strategy is proposed to obtain an adequate low order model. Numerical examples that use the simulated data and the proposed HM(n,m,s) modeling approach are also given. The results of the examples reveal that the advantage of the HM(n,m,s) modeling approach is remarkable.
点击下载:
PDF (553KB)
返 回