分层结构方法是目前在复杂系统故障诊断中常用的方法．但由于电路系统中元器件的容错性问题，如直接采用一般的聚类方法，效果并不理想．因此，在构造加性模糊聚类模型的基础上，提出了一种适合电路容错特性的模糊聚类模型匹配算法．将各故障在特征空间中的容错区间作为集合，用Ｈａｕｓｄｏｒｆ距离测定集合间的相似程度来决定类别．用一个实用电路作为例子，采用本算法对故障集进行了聚类分析，并与模糊Ｃ－划分法进行了对比分析．结果表明，该算法是有效的 Hierarchical structure method is commonly used in complex system fault diagnosis. However, due to the fault tolerance of the components in the circuit system, such as directly using the general clustering method, the effect is not satisfactory. Therefore, on the basis of constructing additive fuzzy clustering model, a fuzzy clustering model matching algorithm suitable for circuit fault tolerance is proposed. The fault tolerance interval in each feature space is taken as a set, and the Hausdorff distance is used to determine the similarity between the sets. Using a practical circuit as an example, this algorithm is used to cluster the fault set and compare it with the fuzzy C-partitioning method. The result shows that this algorithm is effective