Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (36): 144-146.DOI: 10.3778/j.issn.1002-8331.2010.36.039

• 数据库、信号与信息处理 • Previous Articles     Next Articles

Evidence theory information fusion algorithm based on fuzzy reasoning

WANG Yun-fei,LI Hui,LI Yun-bin   

  1. School of Electronic Information,Northwestern Polytechnical University,Xi’an 710129,China
  • Received:2010-06-22 Revised:2010-08-30 Online:2010-12-21 Published:2010-12-21
  • Contact: WANG Yun-fei


王云飞,李 辉,李云彬   

  1. 西北工业大学 电子信息学院,西安 710129
  • 通讯作者: 王云飞

Abstract: Evidence theory has a relatively strong theoretical basis which can deal with randomness or ambiguity caused by uncertainty.However,application of evidence theory is difficult to determine the mass function.To solve this problem,this paper proposes an evidence theory information fusion algorithm based on fuzzy set.The method uses fuzzy theory in the Gaussian fuzzy membership function to obtain a probable characteristic under observation likelihood function,and the resulting likelihood function gets the credibility of the information provided by the sensors.Then the reliability of each sensor is changed to a mass function.Finally,multi-sensor information is combined by using evidence theory.Simulation of target recognition shows that the results obtained have higher accuracy and reliability.

Key words: fuzzy information, evidence theory, information fusion, membership function

摘要: 证据理论具有比较强的理论基础,能处理随机性或模糊性所导致的不确定性。但证据理论应用中基本概率分配函数(mass函数)难以确定,针对这一问题,提出了一种基于模糊推理的证据理论信息融合算法。该方法利用模糊理论中的高斯隶属度函数来获得模糊观测下具有概率特性的似然函数,并且由此似然函数得到每个传感器提供信息的可信度;再将各传感器的可信度转化成基本概率赋值函数即mass函数;最后利用证据理论对多传感器信息进行融合。对目标识别的仿真试验表明该方法获得的结果比直接结果具有更高的精度和可靠性。

关键词: 模糊信息, 证据理论, 信息融合, 隶属函数

CLC Number: