This study is concerned with the control problem of networked non-linear systems with communication delay and data dropout. A networked non-linear predictive control scheme is proposed to compensate for communication delay and data dropout actively rather than passively. The stability analysis shows that the stability problem of the closed-loop networked non-linear predictive control system is converted to the one of a conventional non-linear control system without network. The simulations illustrate that the proposed scheme can compensate for communication delay and data dropout, and achieve the same control performance of the local closed-loop control system without network.
IntroductionWith the rapid development of communication, control and computer technology, the networked control system (NCS) is currently playing a very important role in the Internet of things. The NCS is a spatially distributed control systems via communication networks. Its structure is much more complex than traditional control systems. Although the NCS compared with conventional control systems has many advantages, there also exist various challenging issues, for example, transmission delay, packet loss, network scheduling, communication constraints, signal quantification, data security and so on [1, 2]. To make a networked system have good control performance, a great amount of research work has been carried out in system modelling, controller design, stability analysis and implementation of NCSs in recent years [3][4][5][6][7]. It can be classified as three main research directions: control over network, control of network and control of Internet of things. The control over network uses control strategies to compensate for network delay, packet loss and communication constraints to make the networked system achieve desired control performance [8][9][10][11]. The control of network employs control strategies to optimise network routing and information scheduling so that the service of the entire network is of high quality [12,13]. The control of Internet of things combines both control strategies and scheduling strategies to achieve efficient control of things via networks [14,15].In recent years, most of research in NCSs focuses on the control over network. In particular, to actively compensate for network delay and data loss and achieve desired control performance, the networked predictive control scheme for networked linear control systems was proposed in [16,17]. The design, analysis and implementation of networked predictive control systems have widely been studied in recent years [5,18]. The robustness and stability criteria of networked predictive control systems have been derived for NCSs with internal and external uncertainties [8,19]. The networked predictive control scheme has also extended to some particular non-linear control systems, for example, output feedback linearisable systems [20], Hammerstein systems [21], Wiener systems [22] and constrained systems [23].Although some specific types of networked non-linear contro...