针对区间信念结构组合问题，考虑群决策环境下专家意见冲突的情形，提出一种冲突的区间信念结构下基于证据推理规则的群决策方法。首先，依据决策者 (或专家) 对备选方案的评价信息形成群体决策问题，将专家的评价信息 (观点) 作为证据，检查评价信息的有效性并做归一化处理; 其次，依据证据支持度确定专家权重向量，并运用证据推理规则融合所有评价信息; 最后，运用最小最大后悔值法选出最优方案。实例分析表明，文中所提出的方法在证据组合过程中的证据融合结果始终保持一致，组合结果合理、收敛，而且能保持证据的特异性。

A group decision making method based on evidential reasoning rule under conflicting interval belief structures was proposed to solve the problem of interval belief structures combination，considering the conflict expert opinions in the group decision-making environment． Firstly，group decision-making problems were formed based on the evaluation information of decision makers (or experts) on alternatives． Expert evaluation information (points of view) was taken as evidence，and the validity of evaluation information was checked and normalized． Secondly，the expert weight vector was determined by the support of the evidence，and all the evaluation information was fused with the evidential reasoning rule． Finally，the minimum and maximum regret value method was used to select the optimal scheme． The case analysis shows that the proposed method has consistent evidence fusion results in the process of evidence combination，and the combination results are reasonable and convergent． In addition，the spe-cificity of the evidence can also be maintained．