A novel self-triggered sampling scheme in networked control systems

This paper proposes a novel self-triggered sampling scheme for the execution of sampling in networked control systems by taking into consideration network-induced delays and data packet dropouts. Using this scheme, the next sampling period is dynamically obtained with respect to (a) the desired performance; (b) the latest accepted time-stamped control packet; and (c) the allowable communication delay and the maximum allowable number of successive data dropouts. This scheme can adaptively adjust the sampling period to reduce communication loads with maintaining the desired control performance. Compared with some existing ones, this scheme does not require continuous measurement of the system state and on-line estimation of a triggering condition. An inverted pendulum is employed to demonstrate the effectiveness of the proposed scheme.