基于分离BDD的通用多阶段任务系统可靠性分析

静态的通用多阶段任务系统的不完全覆盖、跨阶段依赖性极大地增加系统可靠性分析的复杂性和难度.基于SEA思想给出两级分离的组合方法将任务和阶段的不完全覆盖分割, 通过新增的阶段代数规则和向后阶段依赖运算式处理跨阶段依赖性,并且充分利用高效的多阶段系统双值决策图,实现模型规模和运算的简化.最后,通过案例分析, 将此分离BDD方法与三值决策图和多状态双值决策图对比, 验证其精确、高效和模块化的优点.

Analysis of generalized phased-mission systems reliability based on separable binary decision diagram

Accounting for static generalized phased-mission systems, imperfect coverage and dependencies across phases increase complexity and difficulty of system reliability analysis obviously. Based on SEA, we present a two-level separable combinatorial method which divides mission and phase imperfect coverage. And then we add some new rules of phase algebra and a new phase dependence operation to ordinary rules and operations, which deal with the dependencies across the phases. The effective PMS-BDD approach is also synthesized in our approach to decrease the model size and computational complexity. In this paper, the approach is compared with several existing approaches by a case study, such as a family of decision diagrams called TDD and MBDD, in order to illustrate its precise, efficient and modular advantages.