论文部分内容阅读
由于车联网是物联网在现代城市交通网络中的具体应用,车联网的数据融合已成为物联网信息感知领域中一个重要的研究课题.针对传统DS证据理论在证据合成过程中存在冲突证据分配不合理、融合方法收敛效果差以及多BBM的证据推理等问题,提出了一种基于等距映射的证据推理方法(isometric mapping evidential reasoning,IMER).IMER,方法根据多BBM证据体间相似关系求取每一个证据体的低维嵌入向量,并计算出低维证据体相似度,以实现对车联网中多BBM证据体的证据推理.实验结果表明本文方法可以合理地分配冲突证据,同时具有较好的收敛性和有效性.
As car networking is a concrete application of IOT in modern urban transportation network, the data integration of car networking has become an important research topic in the field of information perception of Internet of Things. In view of the fact that there is conflict evidence distribution in the process of evidence synthesis in traditional DS evidence theory Reasoning, poor convergence of fusion methods and evidence reasoning of multi-BBM, an isometric mapping evidential reasoning (IMER) is proposed.IMER is based on the similarity between multiple BBM evidences And the low-dimensional embedded vectors of each evidence body, and calculates the similarity of low-dimensional evidence body in order to realize the evidence reasoning of multi-BBM evidence body in the car networking.The experimental results show that this method can reasonably distribute the evidence of conflict and has better Convergence and effectiveness.