Sliding Mode Observer Design for Discrete Nonlinear Time-delay Systems with Stochastic Communication Protocol

• Published in: International journal of control, automation, and systems Vol. 17; no. 7; pp. 1666 - 1676
• Main Authors: Chen, Shuai, Guo, Jun, Ma, Lifeng
• Format: Journal Article
• Language: English
• Published: Bucheon / Seoul Institute of Control, Robotics and Systems and The Korean Institute of Electrical Engineers 01.07.2019; Springer Nature B.V; 제어·로봇·시스템학회
• Subjects: Aircraft; Aircraft control; Algorithms; CCD cameras; Communication; Communication channels; Communications systems; Control; Control stability; Control systems design; Data transmission; Disturbance observers; Dwell time; Engineering; Inspection; Markov chains; Mechatronics; Motion control; Motion stability; Nonlinear systems; Protocol; Robot dynamics; Robotics; Robots; Sliding mode control; Stability analysis; Switching; Time delay systems; Tracking errors; Trajectory analysis; Twisting; Ultrasonic testing; 제어계측공학; Markovian jump systems; stochastic communication protocol; sliding mode observer; networked systems
• DOI: 10.1007/s12555-018-0727-0
• ISSN: 1598-6446

In this paper, the network-based sliding mode observer is investigated for a class of discrete nonlinear time-delay systems with stochastic communication protocol. The stochastic communication protocol is adopted to regulate the transmission order of the measurements from multiple sensor nodes, which could effectively avoid data collisions. Under the scheduling of communication protocol, only one sensor node is allowed to get access to the shared communication network at each time step for data transmission. Moreover, the stochastic communication protocol is governed by a Markov chain, which converts the protocol-constrained system into a Markovian jump system. It is the purpose to design a sliding mode observer such that, with the stochastic communication protocol, the trajectories of the estimation error system are driven into a band of the sliding surface and, in subsequent time, the sliding motion is mean-square asymptotically stable. By solving a minimization problem, the sufficient conditions for the desired sliding mode observer are established. Finally, an illustrative example is given to demonstrate the effectiveness of the proposed algorithm.