Design of an extended kalman filter for UAV localization

Guoqiang Mao*, Sam Drake, Brian D.O. Anderson

*Corresponding author for this work

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    120 Citations (Scopus)

    Abstract

    Unmanned aerial vehicles (UAVs) are increasingly used in military and scientific research. Some miniaturized UAVs rely entirely on the global positioning system (GPS) for navigation. GPS is vulnerable to accidental or deliberate interference that can cause it to fail. It is not unusual, even in a benign environment, for a GPS outage to occur for periods of seconds to minutes. For UAVs relying solely on GPS for navigation such an event can be catastrophic. This paper proposes an extended Kalman filter approach to estimate the location of a UAV when its GPS connection is lost, using inter-UAV distance measurements. The results from a recent trial conducted by DSTO in Australia with three UAVs are presented. It is shown that the location of a manoeuvering UAV that has lost the GPS signal can be determined to an accuracy of within 40m of its true location simply by measuring the range to two other UAVs at known location, where the range measurement error has a zero mean and a standard deviation of 10m.

    Original languageEnglish
    Title of host publicationConference Proceedings of 2007 Information, Decision and Control, IDC
    Pages224-229
    Number of pages6
    DOIs
    Publication statusPublished - 2007
    Event2007 Information, Decision and Control, IDC - Adelaide, Australia
    Duration: 12 Feb 200714 Feb 2007

    Publication series

    NameConference Proceedings of 2007 Information, Decision and Control, IDC

    Conference

    Conference2007 Information, Decision and Control, IDC
    Country/TerritoryAustralia
    CityAdelaide
    Period12/02/0714/02/07

    Fingerprint

    Dive into the research topics of 'Design of an extended kalman filter for UAV localization'. Together they form a unique fingerprint.

    Cite this