基于模煳集合的证据理论信息融合方法.pdfVIP

第25 卷第3 期 控 制 与 决 策 2010 年3 月 Vol. 25 No . 3 Cont rol and D e cisi on Mar. 20 10 : 100 10920( 2010) 韩 峰, 杨万海, 袁晓光 ( , 710071) : 针对证据理论应用中基本概率分配函数( mass 函数) 和多传感器信息融合中 传感器测量数据的可靠程度 均难以确定的问题, 提出了 一种基于模糊集合的证据理论信息融合方法. 该方法首先利用模糊理论中的相关性函数 来计算多传感器的相互支持程度; 然后由隶属函数得到每 个传感器提供信息的可信度; 再将 传感器的支持度和可 信度转化成基本概率分配函数即mass 函数; 最后利用证据理论对多传感器信息进行融合. 仿真结果表明, 该方法获 得的结果具有更高的精度和可靠性. : 多传感器; 证据理论; 模糊集合; 信息融合 : TP2 12. 9 : A Evidence theory information fusion method based on fuzzy set H A N Fe ng , YA N G Wanha i, Y UA N X i aog uang ( S hool of Ele t roni Engineering , Xidian University , Xi an 710071, China. Correspondent: H A N Feng , Email: hanfeng@ xat u. edu. n) Abstract: Fo used on the problem that it is diffi ult to set up the basi probability assignment fun tion ( mass fun t ion) in the eviden e t heory and determine the reliability of ea h sensor in the pro ess of the multi sensors data fusion, an information fusion method based on fuzzy set and eviden e theory is proposed. The mutual supportability of mult iple sensors is obtained from the orrelat ion fun tion. Then by using the membership fun t ion, the reliability of information provide by ea h sensor is gained. T he supposed fusion result an be produ ed on the basis of eviden e theory. T he method is simple omput ationally and an obje tively refle t the reliabilit y of ea h sensor and interrelationship betw een these sensors. Finally, the simulation result shows that the fusion results have higher pre ision and reliability ompared w ith other methods. Key words : Mult i sensor ; Eviden e theory ; Fuzzy set; Information fusion 1 引 言 ; DS .