Abstract:

Evidence theory is widely used in both temporal and spatial domain. Since evidence fusion in time domain is sequential and dynamic, methods of dealing with evidence conflict in space domain may not be suitable in time domain. When belief functions obtained in different time nodes are combined, it is significant to determine the dynamic credibility of evidence in time domain. The Markovian requirement of temporal fusion is analyzed based on the Dempster’s combination rule and the evidence discount theory. And the model of credibility attenuation is established to get the dynamic evidence credibility. Then the evidence is discounted by the dynamic discount factor. Numerical examples demonstrate that such a model can satisfy the requirement of data fusion in time domain. By managing conflicts between belief functions better, this approach is helpful for making proper and solid decisions.