Embedded estimator in predictive feedback control

L. Giovanini

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

In this work, a new approach to design predictive feedback control for SISO systems is presented. The proposed formulation relies on the development of a single step predictor based on an autoregressive moving average with external input (ARMAX) model. Although no explicit observer is actually involved in the implementation, this predictor implicitly includes one since the input-output model subsumes an observer. Exploiting this idea the resulting ARMAX model is extended to include extra outputs to improve the quality of the prediction for systems with large time delay and nonmeasurable disturbances. The resulting predictor is used to develop a predictive feedback controller. This new formulation of predictive feedback control includes feedback and feedforward actions. Simulations of two linear systems illustrate the applicability of the control algorithm.
LanguageEnglish
Pages343-359
Number of pages17
JournalISA Transactions
Volume43
Issue number3
DOIs
Publication statusPublished - 2004

Fingerprint

Predictive Control
feedback control
estimators
Feedback Control
Feedback control
Predictors
Estimator
Observer
predictions
Feedback
Formulation
Autoregressive Moving Average
Output
SISO (control systems)
Feedforward
autoregressive moving average
formulations
Control Algorithm
Linear systems
output

Keywords

  • predictive feedback control
  • estimator
  • dynamic observers
  • feedforward control
  • inferential control

Cite this

Giovanini, L. / Embedded estimator in predictive feedback control. In: ISA Transactions. 2004 ; Vol. 43, No. 3. pp. 343-359.
@article{808aeff6ccdc4ab8b18401e7a25d7e7e,
title = "Embedded estimator in predictive feedback control",
abstract = "In this work, a new approach to design predictive feedback control for SISO systems is presented. The proposed formulation relies on the development of a single step predictor based on an autoregressive moving average with external input (ARMAX) model. Although no explicit observer is actually involved in the implementation, this predictor implicitly includes one since the input-output model subsumes an observer. Exploiting this idea the resulting ARMAX model is extended to include extra outputs to improve the quality of the prediction for systems with large time delay and nonmeasurable disturbances. The resulting predictor is used to develop a predictive feedback controller. This new formulation of predictive feedback control includes feedback and feedforward actions. Simulations of two linear systems illustrate the applicability of the control algorithm.",
keywords = "predictive feedback control , estimator , dynamic observers, feedforward control, inferential control",
author = "L. Giovanini",
year = "2004",
doi = "10.1016/S0019-0578(07)60153-0",
language = "English",
volume = "43",
pages = "343--359",
journal = "ISA Transactions",
issn = "0019-0578",
number = "3",

}

Embedded estimator in predictive feedback control. / Giovanini, L.

In: ISA Transactions, Vol. 43, No. 3, 2004, p. 343-359.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Embedded estimator in predictive feedback control

AU - Giovanini, L.

PY - 2004

Y1 - 2004

N2 - In this work, a new approach to design predictive feedback control for SISO systems is presented. The proposed formulation relies on the development of a single step predictor based on an autoregressive moving average with external input (ARMAX) model. Although no explicit observer is actually involved in the implementation, this predictor implicitly includes one since the input-output model subsumes an observer. Exploiting this idea the resulting ARMAX model is extended to include extra outputs to improve the quality of the prediction for systems with large time delay and nonmeasurable disturbances. The resulting predictor is used to develop a predictive feedback controller. This new formulation of predictive feedback control includes feedback and feedforward actions. Simulations of two linear systems illustrate the applicability of the control algorithm.

AB - In this work, a new approach to design predictive feedback control for SISO systems is presented. The proposed formulation relies on the development of a single step predictor based on an autoregressive moving average with external input (ARMAX) model. Although no explicit observer is actually involved in the implementation, this predictor implicitly includes one since the input-output model subsumes an observer. Exploiting this idea the resulting ARMAX model is extended to include extra outputs to improve the quality of the prediction for systems with large time delay and nonmeasurable disturbances. The resulting predictor is used to develop a predictive feedback controller. This new formulation of predictive feedback control includes feedback and feedforward actions. Simulations of two linear systems illustrate the applicability of the control algorithm.

KW - predictive feedback control

KW - estimator

KW - dynamic observers

KW - feedforward control

KW - inferential control

U2 - 10.1016/S0019-0578(07)60153-0

DO - 10.1016/S0019-0578(07)60153-0

M3 - Article

VL - 43

SP - 343

EP - 359

JO - ISA Transactions

T2 - ISA Transactions

JF - ISA Transactions

SN - 0019-0578

IS - 3

ER -