一种基于决策距离测量与D-S证据理论结合的多源冲突证据信息融合算法

An Information Fusion Algorithm for Multi-source Conflict Evidence Based on Decision Distance Measurement and D-S Evidence Theory

应用D-S证据理论能够有效的解决多传感器的信息融合问题,而在证据间冲突情况严重时,应用D-S合成规则可能获得有悖常理的结论,针对这一问题,通过细致研究多种典型处理冲突证据信息融合的方法,引入决策距离测量和证据间距离函数概念,应用证据间支持度、可信度和平均信任度函数等处理证据的方法,将多源冲突证据重构数学模型合理化处理,再应用D-S证据合成规则进行多源信息融合,得到一种基于决策距离与D-S证据理论结合的多源冲突证据信息融合算法.由仿真结果分析可知,该算法能解决D-S合成规则融合冲突证据信息存在的问题,得出比较合理的信息融合结论.

The application of D-S evidence theory can effectively solve the problem of multi-sensor information fusion.But in the case of highly conflicting evidence,the application of D-S synthesis rules may obtain unconventional conclusions.To solve this problem,through careful study of various methods for dealing with conflict evidence information fusion,the concept of decision distance measurement and the distance function between evidences is introduced,and the method of processing evidence,such as support degree,credibility and average trust function,is applied to rationalize the mathematical model of multi-source conflict evidence reconstruction.Then,multi-source information fusion is conducted by applying the D-S evidence synthesis rule,and a multi-source conflict evidence fusion algorithm based on decision distance and D-S evidence theory is obtained.It is known from the simulation results that the algorithm can solve the problem of the fusion of conflict evidence information in D-S synthesis rules and obtain a reasonable information fusion conclusion.