Event-triggered distributed H∞ state estimation with packet dropouts through sensor networks

Abstract

This study is concerned with the event-triggered distributed H∞ state estimation problem for a class of discrete-time stochastic non-linear systems with packet dropouts in a sensor network. An event-triggered communication mechanism is adopted over the sensor network with hope to reduce the communication burden and the energy consumption, where the measurements on each sensor are transmitted only when a certain triggering condition is violated. Furthermore, a novel distributed state estimator is designed where the available innovations are not only from the individual sensor, but also from its neighbouring ones according to the given topology. The purpose of the problem under consideration is to design a set of distributed state estimators such that the dynamics of estimation errors is exponentially mean-square stable and also the prespecified H∞ disturbance rejection attenuation level is guaranteed. By utilising the property of the Kronecker product and the stochastic analysis approaches, sufficient conditions are established under which the addressed state estimation problem is recast as a convex optimisation one that can be easily solved via available software packages. Finally, a simulation example is utilised to illustrate the usefulness of the proposed design scheme of event-triggered distributed state estimators.