As an important parameter in the evidential reasoning (ER) rule, evidence weight has a significant impact on the fusion results. In the existing research, much attentions have been paid on calculating or obtaining weights under different contexts, while the research on the influence and the feasible range of evidence weight is relatively insufficient. In this article, to reveal the influence of evidence weight and provide guidance for determining acceptable weight range, the impact of evidence weight on the fusion results is evaluated by two evaluation indicators, namely significant grade and fluctuation coefficient. With the analytical expression of the ER rule, the mapping relationship between weight and fusion results is established in reverse. On the premise of ensuring the effectiveness of fusion results, an optimization model for solving the acceptable weight interval (AWI) is established according to the mapping relationship. The effectiveness of the proposed method is verified by two case studies on both the multiple gyroscopes optimal decision-making and the car selection problem.
Inference Analysis on the Evidential Reasoning Rule Under Evidence Weight Variations
IEEE Transactions on Aerospace and Electronic Systems ; 60 , 1 ; 430-448
2024-02-01
3926628 byte
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Experiments in Evidential Locative Reasoning
British Library Conference Proceedings | 1995
|