在加权模糊聚类算法中,属性权重确定的合理性是一个重要问题.鉴于用区间数描述决策者推理模糊性的优越性,提出属性权重用区间数表示,由区间层次分析法获得属性对聚类的贡献度,并以该区间为约束条件,提出了可同时获得属性权重和聚类结果的模糊C均值聚类新算法.实验结果表明,该算法以决策者的经验和偏好为监督,可避免迭代计算陷入不必要的局部极小解,能够提高权重分配的合理性,进而得到了更为准确的聚类结果. In the weighted fuzzy clustering algorithm, the rationality of attribute weight determination is an important issue. In view of the superiority of decision-makers’ reasoning ambiguity using interval numbers, we propose that attribute weight is expressed by interval number, This paper proposes a new fuzzy C-means clustering algorithm that can simultaneously obtain attribute weights and clustering results by using the interval as constraint condition.The experimental results show that the algorithm is supervised by the decision makers’ experience and preferences, Avoiding iterative computation into unnecessary local minimum solution, improving the rationality of weight distribution and getting more accurate clustering results.