Output feedback based robust iterative learning control via a heuristic approach for batch processes with time-varying state delays and uncertainties. (August 2022)

• Record Type: Journal Article
• Authors: Wei, Junyu; Tao, Hongfeng; Hao, Shoulin; Paszke, Wojciech; Gałkowski, Krzysztof
• Is Part Of: Journal of process control. Volume 116(2022)
• Page Start: 159
• Page End: 171
• Publication Date: 2022-08
• Subjects: Batch processes -- Iterative learning control -- Time-delay -- Repetitive processes -- Output feedback -- Two-stage heuristic approach; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2022.06.008
• Languages: English
• ISSNs: 0959-1524

Abstract

Abstract: This paper proposes an output feedback (OF) based PD-type robust iterative learning control (ILC) scheme for a class of discrete batch processes with time-varying delays and uncertainties in the absence of accurate state measurements, which is convenient for practical applications. Based on a repetitive process model and Lyapunov–Krasovskii theory, sufficient conditions are established in terms of linear matrix inequalities (LMIs) to ensure the robust stability along the trial for all admissible uncertainties. A two-stage heuristic approach is developed to iteratively calculate the feasible ILC gains based on a pre-designed state feedback (SF) controller, which circumvents the non-convex stability conditions expressed in bilinear matrix inequalities caused by OF. In addition, the conditions can be extended to the case of non-repetitive uncertainties in the system and meet the prescribed H ∞ performance level. Finally, a simulation of injection molding process is given to demonstrate the effectiveness of the proposed method. Highlights: PD-type iterative learning control (ILC) for batch processes with time-varying delays. The robust H ∞ performance index is considered under non-repetitive uncertainties. Robust ILC updating law based on linear repetitive process model setting. Using output information is beneficial to practical implementation. The heuristic algorithm circumvents the non-convex stability conditions.