针对信息融合中冲突证据的组合问题,提出一种新的利用均衡信度分配准则的冲突证据组合方案.首先分析原始Dempster组合规则及其改进组合规则的不足,然后利用均衡信度分配准则合成新的mass函数.数值实验表明,新组合方法有效地解决了现有Dempster证据组合规则中存在的Zadeh悖论、一票否决、鲁棒性和公平性等问题.
To solve the combination problem of conflict evidence in information fusion, a combination approach of conflict evidence based on proportional belief assignment criterion is presented. Firstly, the drawbacks of original Dempster' s rule of combination as well as several alternative combination approaches proposed so far are analyzed. Secondly, a new mass function based on proportional belief assignment criterion is combined. Finally, the results of numerical examples show that the new combination method can efficiently solve the Zadeh counterintuitive contradiction, the loss of majority opinion, robustness and the fairness of original and improved Dempster's rules of combination.