On Two Classes Of Networked Control Systems

In this paper, two classes of networked control systems are studied. The first one is an observer-based synchronization of a continuous chaotic system under information constraints. For a general chaotic master system and its observer-based slave system, which are connected by a limited capacity channel, a practical quantized scheme is designed such that the synchronization error is ISS on the transmission error, while the transmission error converges to zero exponentially. Therefore, the synchronization error converges to zero asymptotically under information constraints. The obtained result overcomes difficulties of previous works in which the synchronization error can't converge to zero under information constraints. Finally, a simulation example is presented to illustrate the result.The second one is a regular and impulsive-free singular networked control system, where the sensor node is time-driven, but the controller and executor are event-driven. Then, such a singular networked control system can be modeled as an asynchronous dynamical system in the case of no networked-induced delay and no data packet dropout. Meanwhile, being regarded a time-varying transmission period as a time-varying uncertain parameter, a sufficient condition on exponential stability is given by the Lyapunov function method and LMI (linear matrix inequality) techniques. Taking advantage of the Matlab LMI toolbox, we can not only prove the stability of the networked system, but also construct a state feedback control law. A simulation example is given to illustrate the relevant result.