Estimating the State of Distributed Dynamical Systems Based on Limited Measurements by means of Cooperative Filtering with Application to ad-hoc Vehicle Network

Abstract

In this paper a decentralized approach to state estimation of distributed dynamical systems over unreliable communication network subject to uncertain and limited measurements is addressed. Specifically, a distributed filtering concept for enhancing the estimation accuracy of vehicles state from uncertain position data in ad-hoc vehicle networks is presented. The filter relies on a variety of position measurements obtained from the on-board vehicle positioning system, from other nearby cooperating vehicles, as well as from the immediate roadside environment via communication. Direct distance measurements between vehicles as well as between vehicles and the static elements of the roadside infrastructure can be used, if available, as highly accurate reference measurements. The paper demonstrates how successfully the idea of cooperative filtering in addressing the demands of both sensor fault tolerance and enhanced estimation accuracy in distributed systems over ad-hoc communication networks can be used.