A Generalized Distance-Based Total Uncertainty Measure in Complex Evidence Theory

In this paper, a generalized total uncertainty (GTU) measure method is presented directly under the complex evidence theory framework, rather than under the probability framework. GTU is defined based on the distance between the belief interval of each singleton and the most uncertain case, which is normalized by the normalization factor. We proved the properties of GTU. There is an average evidence that represents the average acceptable environment for data from sensors. According to the GTU difference of each data and the average evidence, data from different sensors are given different weights. Each weight represents the trustworthiness of each sensor data in the fusion process, and giving a low weight to abnormal data can avoid false conclusions. The practicability and rationality of the method are verified by numerical examples and applications.