论文部分内容阅读
在应用证据理论融合多源信息时,信任函数冲突是面临的一个重要难题.基于闭世界假设,冲突的产生是由于信息源的误判或无知,对不可靠信息源采用折扣系数弱化其作用是一个解决途径,然而折扣系数生成是一个被忽略的问题.为此,本文提出一种基于证据可信度的自适应折扣系数生成方法,根据证据之间的相似性导出合理的证据可信度,并根据该可信度与提供证据的信息源可靠性共同生成折扣系数,信息源的可靠性通过后验对比进行更新.多个实例结果表明,该方法能对多信息源的一致或冲突证据进行有效合成,且合成结果与人们的直觉一致.
Conflict of trust function is an important challenge in the application of evidence theory to fusion of multi-source information.Based on the closed-world hypothesis, the conflict is caused by the misjudgment or ignorance of the source of information and the weakening of the discount coefficient to the unreliable source is However, the generation of discount coefficient is a neglected issue.Therefore, this paper presents a method of generating an adaptive discount coefficient based on the credibility of evidence, derives reasonable evidence credibility according to the similarity between the evidences, Based on this credibility and the reliability of the information sources providing evidence, a discount coefficient is generated, and the reliability of the information source is updated through posterior comparison.The results of several examples show that the method can make consistent or conflicting evidences of multiple information sources Effective synthesis, and the synthesis results and people’s intuition.