Model validation for industrial model predictive control systems

Huang, Biao; Tamayo, Edgar C.

doi:10.1016/s0009-2509(99)00526-6

Cited by 30 publications

(19 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This type of system has been studied for many years in an LMI framework [8] , and represents a flexible tool to design controllers for uncertain systems and nonlinear systems as well.…”

Section: Problem Formulationmentioning

confidence: 99%

Robust model predictive control of continuous uncertain systems*

Liu

Han

2008

J Syst Sci Complex

View full text Add to dashboard Cite

The authors concern robust model predictive control for linear continuous systems with polytopic uncertainties and input constraints. At each sampling time, a piecewise constant control sequence is obtained by solving a set of linear matrix inequalities. The sufficient conditions on the existence of the model predictive control are given, and the robust stability of the closed-loop systems is guaranteed. A simulation example illustrates the efficiency of the proposed method.

show abstract

“…This type of system has been studied for many years in an LMI framework [8] , and represents a flexible tool to design controllers for uncertain systems and nonlinear systems as well.…”

Section: Problem Formulationmentioning

confidence: 99%

Robust model predictive control of continuous uncertain systems*

Liu

Han

2008

J Syst Sci Complex

View full text Add to dashboard Cite

show abstract

“…This makes the performance diagnosis approach different from the set-based approaches (Olaru, De Doná, & Seron, 2008;Seron & De Doná, 2010), which can be used, for example, to detect sensor and actuator failures. In addition, as opposed to the more classical fault detection approaches (Basseville, 1998;Basseville & Benveniste, 1983;Benveniste et al, 1987;Huang & Tamayo, 2000;Jiang, Huang, & Shah, 2009), the diagnosis objective in this paper is not to detect any plant changes, but only those changes that lead to the closed-loop performance degradation (i.e., control-relevant changes). As in Gustafsson and Graebe (1998), Tyler and Morari (1996), the presented diagnosis approach is based on cheap identification of the true system with an economic cost much smaller than that when full re-identification should be performed for performance restoration.…”

Section: Introductionmentioning

confidence: 99%

Least costly closed-loop performance diagnosis and plant re-identification

Mesbah

Bombois

Forgione

et al. 2015

International Journal of Control

View full text Add to dashboard Cite

The inherent time-varying nature of dynamics in chemical processes often limits the lifetime performance of model-based control systems, as the plant and disturbance dynamics change over time. A critical step in the maintenance of model-based controllers is distinguishing control-relevant plant changes from variations in disturbance characteristics. In this paper, prediction error identification is used to evaluate a hypothesis test that detects if the performance drop arises from controlrelevant plant changes. The decision rule is assessed by verifying whether an identified model of the true plant lies outside the set of all plant models that lead to adequate closed-loop performance. A unified experiment design framework is presented in the least costly context (i.e., least intrusion of nominal plant operation) to address the problem of input signal design for performance diagnosis and plant re-identification when the performance drop is due to plant changes. The application of the presented performance diagnosis approach to a (nonlinear) chemical reactor demonstrates the effectiveness of the approach in detecting the cause of an observed closed-loop performance drop based on the designed least costly diagnosis experiment.

show abstract

“…However, one of the key elements to the successful implementation of such control systems is the process model (Huang & Tamayo, 2000), which is used to predict the future behavior of the system output along a prediction horizon that is large enough to encompass all the process dynamics that need to be taken into account in the computation of the control law. Thus, a poor process model will result in biased output predictions, which will affect the control performance and may jeopardize the expected economic benefit associated with the implementation of the control strategy.…”

Section: Introductionmentioning

confidence: 99%