Asynchronous Learning-Based Output Feedback Sliding Mode Control for Semi-Markov Jump Systems: A Descriptor Approach
Zheng Wu,Yiyun Zhao,Fanbiao Li,Tao Yang,Yang Shi,Weihua Gui
DOI: https://doi.org/10.1109/jas.2024.124416
2024-05-28
IEEE/CAA Journal of Automatica Sinica
Abstract:This paper presents an asynchronous output-feedback control strategy of semi-Markovian systems via sliding mode-based learning technique. Compared with most literature results that require exact prior knowledge of system state and mode information, an asynchronous output-feedback sliding surface is adopted in the case of incompletely available state and non-synchronization phenomenon. The holonomic dynamics of the sliding mode are characterized by a descriptor system in which the switching surface is regarded as the fast subsystem and the system dynamics are viewed as the slow subsystem. Based upon the co-occurrence of two subsystems, the sufficient stochastic admissibility criterion of the holonomic dynamics is derived by utilizing the characteristics of cumulative distribution functions. Furthermore, a recursive learning controller is formulated to guarantee the reachability of the sliding manifold and realize the chattering reduction of the asynchronous switching and sliding motion. Finally, the proposed theoretical method is substantiated through two numerical simulations with the practical continuous stirred tank reactor and F-404 aircraft engine model, respectively.
automation & control systems