• DocumentCode
    532333
  • Title

    Conflict problems of D-S evidence in multi-sensor information fusion technology

  • Author

    Chen, Xuhui ; Yang, Yongxu ; Cheng, Huiyan

  • Author_Institution
    Comput. & Commun. Coll., Lanzhou Univ. of Technol., Lanzhou, China
  • Volume
    4
  • fYear
    2010
  • fDate
    22-24 Oct. 2010
  • Abstract
    The D-S reasoning has a big advantage of presentation, resolution and combination for uncertainty information problems, that makes it can be effectively applied in multi-sensor data fusion. However, the Dempster combination rule doesn´t take proper measures to address the reliability, conflict and importance aspects of measuring the source-evidence itself, this weakness may result in the existence contrary to intuition in reasoning process. A synthesis rule, especially the conflict problems produced by the data fusion involved in the D-S reasoning synthesis rule based on the basic ones has been analyzed in this paper. Also, a novel synthesis formula of D-S has been proposed according to the idea of supporting probability of evidence conflict that comes from the average supporting degree of each proposition obtained by using weighted distribution method, and the conception of evidence information and the focal element generated by itself, and it is applied in the multi-sensor information fusion technology, As indicated by the simulation results that the method proposed is a feasible, effective solution.
  • Keywords
    data handling; sensor fusion; D-S evidence conflict problems; Dempster combination; multisensor data fusion; multisensor information fusion technology; source evidence measurement; synthesis rule; uncertainty information problems; Gallium nitride; Mathematical model; Conflict evidence; D-S theory of evidence; Information fusion; Multi-sensor;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Application and System Modeling (ICCASM), 2010 International Conference on
  • Conference_Location
    Taiyuan
  • Print_ISBN
    978-1-4244-7235-2
  • Electronic_ISBN
    978-1-4244-7237-6
  • Type

    conf

  • DOI
    10.1109/ICCASM.2010.5620351
  • Filename
    5620351