Distributed Kalman–Bucy Filter With Embedded Dynamic Averaging Algorithm
- 15 February 2017
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 12 (2) , 1722-1730
Abstract
This paper focuses on distributed estimation using networked sensor agents with local measurements and local communication. A distributed Kalman–Bucy filter implementation is proposed and analyzed for the general case of continuous-time linear time-varying systems as compared to linear time-invariant systems. The sensor agents employ a distributed average tracking algorithm and its extension that accounts for bounded noise to estimate the averages of certain time-varying signals by communicating with their local neighbors in the network. Such estimates are then used to recover certain information in the implementation of the distributed Kalman–Bucy filter. It is shown that using the proposed distributed filter, in the absence of measurement noise, the distributed local estimates approach the centralized filter's estimate asymptotically. In the presence of bounded measurement noise, the distributed local estimates asymptotically approach the centralized filter's estimate within some bound. Simulation results illustrate the good performance of the distributed filter.Keywords
Funding Information
- Deanship of Scientific Research (DSR)
- King Abdulaziz University (18-135-35-HiCi)
This publication has 20 references indexed in Scilit:
- Consensus-Based Linear and Nonlinear FilteringIEEE Transactions on Automatic Control, 2014
- Information Weighted Consensus Filters and Their Application in Distributed Camera NetworksIEEE Transactions on Automatic Control, 2013
- Distributed consensus-based estimation considering network induced delays and dropoutsAutomatica, 2012
- Consensus-based linear distributed filteringAutomatica, 2012
- Coordination of groups of mobile autonomous agents using nearest neighbor rulesIEEE Transactions on Automatic Control, 2003
- General Decentralized Data Fusion with Covariance Intersection (CI)2001
- Data fusion in decentralized sensor networksControl Engineering Practice, 1994
- Fully decentralised algorithm for multisensor Kalman filteringIEE Proceedings D Control Theory and Applications, 1991
- Distributed asynchronous deterministic and stochastic gradient optimization algorithmsIEEE Transactions on Automatic Control, 1986
- Computation and transmission requirements for a decentralized linear-quadratic-Gaussian control problemIEEE Transactions on Automatic Control, 1979