ABSTRACT In recent years, networked control systems have been attracted a lot of attention due to their many applications in various fields. In networked control systems, time delay and data dropout are two negative factors that can reduce system performance or even cause instability. Predictive control is one of the effective methods for compensating the effect of time delay and data dropout in networked control systems. In the proposed predictive control method, the control data is predicted as long as it covers network time delay and data dropout, and the estimation of system state variables is obtained using the Kalman filter observer. After performing control calculations, by sending a packet containing the predicted control data to the operator and selecting the appropriate control data, the desired control performance can be achieved. In this research, the networked control system is considered as a switched system with arbitrary switching signals and based on the theory of switching systems, a criterion for eva‎luating the stability of the closed-loop system with time delay and data dropout is presented. Time delay and data dropout are selected as switching parameters and a subset of dynamic systems in discrete state space is created. In this study, by presenting a common Lyapunov function for all subsets and solving matrix inequalities, the stability of the networked control system is investigated. By using the proposed predictive control method with a state feedback controller, the effects of network time delay and data dropout are well compensated and the desired control performance is achieved. In the proposed predictive method with state feedback controller, the results of the simulation with MATLAB software indicate that the control system is more sensitive facing time delay and data dropout in the forward channel than the feedback channel. Then, a different predictive control method is used by considering the controller dynamics, and in this method, the effects of time delay and data dropout are compensated and the desired control performance is obtained. Also, the comparisons show the superiority of the predictive control method over the non-predictive control method in networked control systems.

مصطفي نصيري