Data-Driven Modelling for a High Order Multivariable Thermal System and Control. Issue 20 (2021)

• Record Type: Journal Article
• Authors: Rodriguez, Carlos; Viola, Jairo; Chen, YangQuan
• Is Part Of: IFAC-PapersOnLine. Volume 54:Issue 20(2021)
• Page Start: 753
• Page End: 758
• Publication Date: 2021
• Subjects: Dynamic Mode Decomposition -- MIMO system identification -- LQR -- ν-gap metric; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2021.11.262
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: The identification of multivariable systems can be challenging due to the complex and coupled dynamics between system inputs and outputs. Thus, traditional system identification like multivariable transfer function may not capture the overall system dynamics relevant for an adequate controller design. This paper presents the use of Dynamic Model Decomposition with Control (DMDc) to identify a model for a high-order multivariable system corresponding to a 16x16 multivariable thermal system with coupled dynamics. The system is controlled using a Linear Quadratic Regulator (LQR) with ν-gap metric controller. Obtained results show that the model obtained using DMDc corresponds to an accurate representation of the system and contributes to performing suitable temperature control.