Dynamic fault tree (DFT) is a powerful modeling approach for reliability analysis of complex system with dynamic failure behaviors. In reality, the tree structure may be highly coupled either by shared basic events or by the high-level dynamic gates. Currently, the application of sequential binary decision diagram (SBDD)-based method for quantitative analysis of such highly coupled DFTs is mainly limited to DFTs whose dynamic gates locate in the bottom of the tree. Moreover, there is no efficient way dealing with the dependencies among different nodes of a SBDD 1-path. This paper makes an improvement to the SBDD-based approach. A generation procedure is proposed to directly construct the binary decision diagram (BDD) model for a DFT with arbitrary tree structure. During the construction, the sequential-dependent information of the tree is derived as several BDD nodes, each indicates a binary-sequential event representing the sequence of two occurred basic events. A topological sorting is applied on each 1-path of the resultant BDD to obtain its contained disjoint cut sequences. Based on this, both qualitative and quantitative analysis can be performed on the DFT with no limitations on tree structure, and its minimal cut sequence set (CSS) is obtained as disjoint. Examples are provided for verification and comparison, and the results illustrate the merits of the proposed approach.

动态故障树（DFT）是一种功能强大的建模方法，用于对具有动态故障行为的复杂系统进行可靠性分析。实际上，树结构可以通过共享的基本事件或高级动态门来高度耦合。当前，基于顺序二进制决策图（SBDD）的方法对这种高度耦合的DFT进行定量分析的应用主要限于动态门位于树底部的DFT。此外，没有有效的方法来处理SBDD 1路径的不同节点之间的依赖性。本文对基于SBDD的方法进行了改进。提出了一种生成程序，用于直接构建具有任意树结构的DFT的二进制决策图（BDD）模型。施工期间 树的顺序相关信息被导出为几个BDD节点，每个节点表示一个二进制顺序事件，表示两个已发生基本事件的顺序。对所得BDD的每个1路径进行拓扑排序，以获得其包含的不相交的剪切序列。基于此，可以对DFT进行定性和定量分析，而对树结构没有任何限制，并且获得的最小切割序列集（CSS）不相交。提供了示例进行验证和比较，结果说明了该方法的优点。可以对DFT进行定性和定量分析，而对树结构没有限制，并且获得的最小切割序列集（CSS）不相交。提供了示例进行验证和比较，结果说明了该方法的优点。可以对DFT进行定性和定量分析，而对树结构没有限制，并且获得的最小切割序列集（CSS）不相交。提供了示例进行验证和比较，结果说明了该方法的优点。