多智能体信息融合(MAIF)系统可以缓解单个专家系统在处理复杂情况时的局限性,因为它允许多智能体协作以解决复杂环境中的问题。D-S证据理论在众多领域有着广泛的应用前景。但是,在处理高度冲突的数据时,传统的D-S组合规则可能会产生与直觉相反的结果。因此,本文提出了一种将重构基本概率分配和信念熵相结合的多智能体系统冲突数据融合方法。首先,使用重构的基本概率分配(BPA)构造每个智能体的初始信任度。然后,通过计算信任熵得到每个证据组的信息量,从而修正可靠性,获得更合理的证据。最后,将最终证据使用Dempster组合规则进行融合,得到结果。数值算例表明了该方法的有效性,提高了MAIF系统辨识过程的精度。

The Multi Agent Information Fusion (MAIF) system can alleviate the limitations of a single expert system when dealing with complex situations, as it allows multi-agent collaboration to solve problems in complex environments. The D-S evidence theory has important applications in fields such as multi-source data fusion and pattern recognition. However, when dealing with highly conflicting data, traditional Dempster combination rules may produce counterintuitive results. Therefore, we propose a conflict data fusion method for multi-agent systems based on reconstructing basic probability assignment and belief entropy. Firstly, use the reconstructed basic probability assignment (BPA) to construct the initial trust level for each agent. Then, by calculating belief entropy, the information content of each evidence group is obtained, thereby correcting reliability and obtaining more reasonable evidence. Finally, the final evidence is fused with the Dempster combination rule to obtain the result. Numerical examples demonstrate the effectiveness of this method and improve the accuracy of the MAIF system identification process.

TP391