A new Takagi-Sugeno-Kang model-based stabilizing explicit MPC formulation: An experimental case study with implementation embedded in a PLC. (30th December 2022)

• Record Type: Journal Article
• Authors: P. G. Mendes, Teófilo; Schnitman, Leizer; dos Reis Nogueira, Idelfonso Bessa; Mafalda Almeida Peixoto Ribeiro, Ana; Egídio Rodrigues, Alírio; Miguel Loureiro, José; Martins, Márcio A.F.
• Is Part Of: Expert systems with applications. Volume 210(2022)
• Publication Date: 2022-12-30
• Subjects: Explicit MPC -- Takagi-Sugeno-Kang Model -- Stabilizing control law -- PLC implementation; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2022.118369
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: A new re-interpretation of explicit MPC solutions via the Takagi-Sugeno-Kang model. A novel stabilizing MPC formulation with relaxed terminal constraints. Application of the proposed MPC embedded in a PLC following IEC-61131-3 standard. Experimental results obtained from a DC motor validate the proposed method. Abstract: This manuscript presents a new fuzzy approach applied to Model Predictive Control (MPC). We propose to re-interpret the table of IF-THEN rules from an explicit MPC solution as an expert system. The first contribution of this work is a modification in the Takagi-Sugeno-Kang (TSK) structure that allows modeling this table of IF-THEN rules without approximation and maintaining the performance of the explicit MPC solution. The second contribution is a new MPC formulation that guarantees closed-loop system stability by combining a one-layer optimization, a suitable Jordan decomposition-based state-space model, and relaxed terminal constraints. These proposals were implemented in a general-purpose Programmable Logic Controller, following the IEC-61131-3 standard, and applied to control the speed of a physical DC motor. The new modified TSK method and the stabilizing MPC formulation showed low-level computational effort and the desired control performance. The results indicate the potential application in standard automation systems available at industrial process plants.