Sub-predictors for network-based control under uncertain large delays

This paper, for the first time, provides sub-predictors for networked control systems (NCSs) under uncertain large communication delays. We use a time-delay approach to NCS and employ sub-predictors to partially compensate large uncertain transmission delays in the sensor-to-controller and controller-to-actuator channels by dividing the long delay into small pieces. We consider systems with norm-bounded uncertainties, and take into account Round-Robin scheduling protocol in sensor-to-controller channel. In comparison with the traditional reduction-based classical predictor involving distributed input, the sub-predictor-based feedback is more friendly in the presence of norm-bounded uncertainties and is simpler for implementation. The sub-predictor-based feedback is further extended to decentralized control of interconnected systems provided that the couplings are not strong. The stability analysis of the closed-loop system is based on the Lyapunov–Krasovskii method and the stability conditions are given in terms of linear matrix inequalities.