DMPC of Networked Systems with Event-Triggered Computation

This chapter investigates the event-triggered DMPC scheme for achieving convergence and constraints satisfaction of NCSs. By applying a trigger in the sensor-to-controller channel in each subsystem, the computation and communication manners are regulated such that the NCS achieves the control goals with less resources consumption compared to the normal DMPC scheme. Specifically, two approaches that contain different event-triggering conditions are reported. In Sect. 2.2, to achieve the consensus property of NCSs with decoupled dynamics, a DMPC algorithm with a cooperative event-triggering condition is presented. The event-triggered condition is designed considering the information received from neighboring subsystems, and the sufficient conditions for global ISS are derived. In Sect. 2.3, a dual-mode DMPC strategy with a decentralized event-triggered computation mechanism is developed for NCSs with bounded uncertainty. The triggering condition is only parameterized on the information of the subsystem itself. Once the states of two interactive subsystems both enter the pre-designed invariant set respectively, no information will be exchanged between them, thus reducing the communicational resources with acceptable closed-loop performance. Examples are provided to verify the effectiveness of the strategies.