A new data fusion algorithm based on the continuous-time decentralized Kalman filter

Yuri A. Vershinin, M. J. West

    Research output: Chapter in Book/Report/Conference proceedingConference proceeding


    Data fusion techniques are used in many tracking and surveillance systems as well as in applications where reliability is of a main concern. The new data fusion algorithm presented allows information from different sensors to be combined in continuous time. Continuous-time decentralized Kalman filters (DKF) are used as data fusion devices on local subsystems. Such a structure gives flexibility for the reconfiguration of a control system. New subsystems can easily be added without needing any redesign of the whole system. The system does not require a central processor and therefore, in the case of failure of some local subsystems (each of which includes a local processor, sensors and actuators) the overall system continues to work. The simulation results show that the performance of the overall system degrades gracefully even if the sensors of some subsystems fail or interconnections are broken. Furthermore, local Kalman filters can effectively reduce subsystem and measurement noise.
    Original languageEnglish
    Title of host publicationTarget Tracking: Algorithms and Applications (Ref. No. 2001/174), IEE
    Pages16/1 - 16/6
    Publication statusPublished - 2001
    EventTarget Tracking: Algorithms and Applications - , Netherlands
    Duration: 16 Oct 200117 Oct 2001


    ConferenceTarget Tracking: Algorithms and Applications

    Bibliographical note

    The full text is currently unavailable on the repository.


    • Kalman filters
    • continuous time filters
    • random noise
    • sensor fusion
    • surveillance
    • tracking
    • continuous-time filter
    • control system reconfiguration
    • data fusion algorithm
    • decentralized Kalman filter
    • local subsystems
    • surveillance systems
    • tracking systems


    Dive into the research topics of 'A new data fusion algorithm based on the continuous-time decentralized Kalman filter'. Together they form a unique fingerprint.

    Cite this