基于随机集理论的多源信息统一表示与建模方法

The Unified Method of Describing and Modeling Multisource Information Based on Random Set Theory

由于信息形式的多样性及其特征的复杂性,使得对不确定、未知性、非精确和不完全等类型信息的表示和建模都要通过相应的方法完成,由于方法的不统一,从而很难实现对异类信息的融合.所以,能否找到一种统一的理论实现多源异类信息的表示与建模,最终实现融合成为信息融合中的关键问题.众多研究者经过多年的探索发现,随机集理论有望解决这个难题.本文首先对各种多源信息进行分类,并介绍几种常用的表示和建模方法及其适用范围;随后引入随机集理论的基本概念和性质,综合论述该理论与已有方法之间的相互转化关系,并进一步论证用随机集统一表示和建模多源信息的可能性;最后,介绍随机集理论在信息融合中的应用并指出未来的发展方向.

Because of the diversity of information types and the complexity of information characteristics, traditionally, the multisource heterogeneous information is described and modeled by the corresponding knew theories, so there is rarely an unified method to fusion them. For several years,researchers have explored the unification of theories enabling the fusion of heterogeneous information and have finally considered random set theory. This paper first classifies available information by qualities （ uncertain, vague, imprecise, etc） and then presents common theories that can be used to cope with them. Random set theory is introduced as a possible framework for unification. The reasons why the individual theories can fit in this framework are detailed. Finally, this paper reviews applications of random set theory in information fusion and discusses the possible directions in the future.