Integrated Design of Event-Triggered Closed-Loop Subspace Predictive Control Scheme

This paper proposes a novel event-triggered closed-loop subspace predictive control algorithm for linear discrete-time systems with an unknown plant model. The integrated observer-based event-triggered system is designed only with the system input and output data. The designed event-triggered law is dependent on the defined input error based on the input-to-state stable Lyapunov (ISS-Lyapunov) function. The receding horizon principle in the typical predictive control is substituted by the event-triggered law, which can considerably reduce the computation load and the time consumption while it maintains its stability and optimality. Moreover, a possible fault detection scheme is also proposed based on the event-triggered law. The simulation results illustrate the effect and the improved performance of the proposed scheme.