Consensus-Based Linear and Nonlinear Filtering

Author

Battistelli, G. ; Chisci, L. ; Mugnai, G. ; Farina, A. ; Graziano, A.

Author_Institution

Dipt. di Ing. dell´Inf., Univ. di Firenze, Florence, Italy

Volume

60

Issue

5

fYear

2015

fDate

May-15

Firstpage

1410

Lastpage

1415

Abstract

This note addresses Distributed State Estimation (DSE) over sensor networks. Two existing consensus approaches for DSE, i.e., consensus on information (CI) and consensus on measurements (CM), are combined to provide a novel class of hybrid consensus filters (named Hybrid CMCI) which enjoy the complementary benefits of CM and CI. Novel theoretical results, limitedly to linear systems, on the guaranteed stability of the Hybrid CMCI filters under collective observability and network connectivity are proved. Finally, the effectiveness of the proposed class of consensus filters is evaluated on a target tracking case study with both linear and nonlinear sensors.

Keywords

filtering theory; linear systems; observability; target tracking; collective observability; consensus-based linear filtering; consensus-based nonlinear filtering; distributed state estimation; hybrid CMCI filters; linear systems; network connectivity; target tracking; Covariance matrices; Equations; Estimation error; Kalman filters; Q measurement; Stability analysis; State estimation; Consensus; distributed state estimation; nonlinear filtering; sensor networks;

fLanguage

English

Journal_Title

Automatic Control, IEEE Transactions on

Publisher

ieee

ISSN

0018-9286

Type

jour

DOI

10.1109/TAC.2014.2357135

Filename

6897960