冲突证据的目标识别系统的研究

目标识别系统中的不确定性信息常常表现为模糊性和冲突性信息.提出了一种在证据理论框架下实现模糊信息融合的方法.针对决策级信息融合中的冲突证据组合问题进行了深入的研究.在对现有组合算法分析的基础上,提出了一种新的冲突证据组合的方法.该方法采用先验概率加权与实际战场环境模糊加权相结合,最后采用归一化的可信度对证据进行加权平均并进行DS融合.仿真实验结果表明,此方法可有效地处理冲突证据,得到合理的融合结果.

Target Recognition System Based on D-S Theory of Conflict Evidence

Evidence theory is widely used in target recognition system. One of the problems in real application is that not only the observation collected by sensors, but also the attributes of targets in the model database may be fuzzy and conflict too. In this situation, how to automatically determine the mass function of conflict information is a difficult issue. In this paper, an intensive research on the combination method of conflict evidence is presented. Based on the analysis of existing modified combination algorithms, a new combination method was proposed, which applied prior probability and fuzzy information of battlefield surroundings to obtain the differentiate weights of multi-sensor. Finally the evidence is averaged by the normalized credit vector and combined n - 1 times by DS rule. Numerical experiment shows that the proposed method can effectively solve the conflict evidence combination problem, and the combination results are rational.