Chattering-free model free adaptive sliding mode control for gas collection process with data dropout. (September 2020)

• Record Type: Journal Article
• Authors: Gao, Xianwen; Weng, Yongpeng
• Is Part Of: Journal of process control. Volume 93(2020)
• Page Start: 1
• Page End: 13
• Publication Date: 2020-09
• Subjects: 00-01 -- 99-00; Data-driven sliding mode control -- Gas collection process -- Predictive control -- Data dropout -- Robustness; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2020.07.003
• Languages: English
• ISSNs: 0959-1524

Abstract

Abstract: Stable pressure control of coke oven gas collectors is difficult due to the problems of nonlinearity, couplings, time-variation, disturbances and data-dropout. This paper proposes a novel chattering-free model free adaptive sliding mode control scheme for the gas collection process of coke ovens. Unlike the conventional data-driven sliding mode control approaches, the proposed controller is based on the data-dropout compensation scheme and predictive control strategy, whereby a novel data-driven sliding surface estimator and a new data-driven sliding mode prediction model are developed to facilitate the controller design. On one hand, the effect of data dropout could be attenuated by applying the proposed estimator. On the other hand, a chattering-free smooth control law with strong robustness to couplings, time-variation and disturbances is obtained owing to the application of the predictive control strategy and the sliding mode control technique. In addition, the convergence of the proposed algorithm could be guaranteed theoretically. Finally, experimental results soundly confirm the efficacy and superiority of the developed approach. Highlights: Data-dropout and model-unavailable are considered for sliding mode surface design. Data-driven sliding mode predictive model is further established. Complex unknown dynamics and data dropout are attenuated by the CMFASMC law. Theory analysis and experimental results are provided for performance evaluation.