Modular solution of dynamic multiple-phased systems  

动态多阶段系统模块化分析(英文)

作　　者：莫毓昌[1,2] 杨全胜[1] 

机构地区：[1]东南大学计算机科学与工程学院,南京211189 [2]浙江师范大学数理信息学院,金华321004

出　　处：《Journal of Southeast University(English Edition)》2009年第3期316-319,共4页东南大学学报（英文版）

基　　金：The National Natural Science Foundation of China(No.60903011);the Natural Science Foundation of Jiangsu Province(No.BK2009267)

摘　　要：A new modular solution to the state explosion problem caused by the Markov-based modular solution of dynamic multiple-phased systems is proposed. First, the solution makes full use of the static parts of dynamic multiple-phased systems and constructs cross-phase dynamic modules by combining the dynamic modules of phase fault trees. Secondly, the system binary decision diagram （BDD） from a modularized multiple- phased system （MPS）is generated by using variable ordering and BDD operations. The computational formulations of the BDD node event probability are derived for various node links and the system reliability results are figured out. Finally, a hypothetical multiple-phased system is given to demonstrate the advantages of the dynamic modular solution when the Markov state space and the size of the system BDD are reduced.针对Markov方法在动态多阶段系统模块化分析中的状态爆炸问题,给出了一种新的动态模块化方法.该方法首先从充分利用动态多阶段系统所包含的静态特性的角度出发,利用集合并操作将阶段故障树的动态模块构造成跨阶段的动态模块.然后,利用变量排序和BDD操作生成模块化MPS对应的系统BDD.针对系统BDD中各种节点连结关系,推导出BDD节点事件概率计算公式,从而计算出动态多阶段系统的可靠度.最后,通过实例说明了动态模块化方法在Markov状态空间和系统BDD尺度均减少时的优越性.

关 键 词：binary decision diagram （BDD） dynamic fault tree Markov chain modular solution 

分 类 号：TP320[自动化与计算机技术—计算机系统结构]