故障网络的连通性能分析和故障源的诊断与识别，一直是互连网络领域重要的研究内容。网络的连通性能和故障诊断能力与故障类型和故障分布息息相关。然而，在以往网络连通性能的研究中，往往只考虑故障类型，忽略故障分布这一因素。而关于故障诊断的相关研究中，所考虑的故障类型又与实际不符，因而在实际应用中局限性很强。因此，本项目拟综合考虑故障类型和故障分布两个因素，关于网络的连通性能和故障诊断展开研究。具体研究内容如下：（1）在度量一些经典网络结构连通能力的基础上，刻画其基数最小的结构割集，明晰其故障分布，进而研究点传递图的结构割集分布。（2）在故障网络无孤立点存在的前提下，关于故障结构的分布提出合理的假设，研究一些经典网络的条件结构连通能力。（3）引入结构故障类型，研究一些经典网络的结构故障诊断度，以及一般图的结构故障诊断度和结构连通度的关系。（4）设计针对一般图和经典网络的结构故障诊断算法，定位故障结构。

The connectedness analysis and diagnosis about fault sets of faulty networks are important research areas in interconnection networks. The connectedness and fault diagnosis ability of networks are closely related to fault types and fault distribution. However, in the past studies about connectedness of networks, researchers only considered the fault types, but ignored the fault distribution. Similarly, since the fault types are inconsistent with real situation, fault diagnosis of related studies have strong limitation in application. Therefore, this project intends to study the connectedness and fault diagnosis of networks under consideration the fault types and fault distribution. We plan to explore the following problems: (1) Based on measuring the structure connectivities of some classical networks, we will characterize the minimum structure cut sets of them and vertex transitive graphs. (2) On the hypothesis that there are no isolated vertices in the faulty networks, we will first assume reasonable distribution of fault structures, and then study the conditional structure connectivities of some classical networks. (3) We will first introduce structure fault type, and then investigate the structure fault diagnosabilities of some classical networks, and explore the relationship between structure fault diagnosability and structure connectivity of graph. (4) We will design the diagnosis algorithms of classical networks and general graphs to locate the fault structures.