中文摘要：

针对传统的相对熵集结模型在赋权方面存在的缺陷，提出了一种基于灰色关联分析的REM。首先应用数值代数中的幂法得到理想的评分向量，再应用灰色关联分析计算理想评分向量与各个专家的评分向量之间的关联度，以此作为专家权重系数，最后将专家权重应用于相对熵集结模型中，来集结专家群组的一致性。通过对嵌入式计算机的性能评价结果表明该方法是有效的，而且可靠性得到很大的提高。

英文摘要：

In view of the irrationality of traditional REM assembly in assigning weights, which can result in wrong evaluation results in a group decision-making, we present the REM assembly based on the grey correlation analysis, which we believe is effective. Firstly, we constructed the ideal score vector by the power method, and the weights of each decision-maker are calculated by the grey correlation degree between the decision results of each individual and the ideal score vector. Finally, we apply weights to relative entropy assembly model to assemble group consistency. A numerical example is given in section 2 of the full paper; the simulation results and their analysis show preliminarily the effectiveness of this algorithm.