Output feedback H∞ control for discrete time singularly perturbed systems with Markov lossy network: The round-robin-like protocol case

• Published in: Applied mathematics and computation 2024-02, Vol.462, p.128338, Article 128338
• Main Authors: Hu, Yue, Kwon, Oh-Min, Cai, Chenxiao, Kim, Yeong-Jae
• Format: Article
• Language: English
• Subjects: Markov lossy network; Networked control; Output feedback control; Round-Robin-like protocol (RRLP); Singular perturbation parameter
• ISSN: 0096-3003; 1873-5649
• DOI: 10.1016/j.amc.2023.128338

In this paper, the output feedback H∞ control is investigated for the networked singularly perturbed systems (SPSs) with Markov lossy network, in which the packet loss probability varies in different modes. A new network mode-dependent round-robin-like protocol (RRLP) is introduced to the networked SPSs, which can dynamically adjust the number of the selected sensor nodes according to the network environment condition. Based on this RRLP and Markov lossy network, a Markov mode and scheduling order co-dependent output feedback controller is constructed to ensure the system control performance. By constructing a novel Lyapunov-Krasovskii functional dependent on the singular perturbation parameter, Markov mode, and scheduling order simultaneously, sufficient conditions of mean square stability for SPSs are obtained, and numerical stiffness is avoided. Finally, a direct current motor-controlled inverted pendulum example is proposed to verify the effectiveness of our main results. •Compared with the conventional RRP, a network mode-dependent RRLP is first introduced into SPSs, by which the number of scheduled sensors can be adjusted appropriately according to different network environments.•Based on the Markov lossy network and multi-node RRP, a novel output feedback controller is designed, which can effectively improve the control performance of system.•By constructing a new Lyapunov-Krasovskii functional dependent on SPP, Markov mode, and scheduling order simultaneously, the sufficient condition of the mean square stability is obtained for SPSs.