Discrete-Time Periodic Event-Triggered Distributed Set-Membership Estimation Over Sensor Networks

Abstract

This paper investigates the event-triggered distributed set-membership estimation problem over sensor networks. Firstly, a novel discrete-time periodic event-triggered mechanism (ETM) is originally proposed to decide when the sensor node transfers the measurement output to its neighbour nodes. Compared with the existing ETMs, the proposed approach only monitors the measurements at each periodic sampling point so that it reduces the energy consumption caused by detection and calculation. Secondly, a piecewise auxiliary function is introduced to deal with the sawtooth characteristic of periodic sampling signals. With the help of such auxiliary function, a co-design criterion of the set-membership estimator and ETM is proposed under the unknown-but-bounded (UBB) noise. Finally, a recursive optimization algorithm is proposed to obtain the optimized ellipsoid estimation set. A classic numerical example is provided to validate the effectiveness of the developed approach.