摘要
对于可修复系统,其状态数随系统规模呈指数增长。文中在系统元件类型相同的假设条件下,以维修顺序为突破口剖析马尔可夫模型,证明在某几种维修顺序条件下,可由类状态概率来计算类状态中单个状态的概率;并以可修复网络系统为例提出了从类状态到不交和状态的映射算法,从而使计算网络可用度的复杂度大大降低。仿真结果表明合并策略和映射算法的正确。
For repairable system, its state number grows exponentially with the system scale. Firstly, under the assumptions of same type of system components, this paper analyzes Markov model using the maintenance sequence as a breakthrough point, and proves that for certain maintenance sequences, the single state probability can be calculated according to the class state probability. What is more, this paper puts forward a mapping algorithm from class state to SDP state by taking the repairable network as example, so that the calculation complexity of the availability of network can be reduced greatly. Finally the correctness of the merger strategy and the mapping algorithm are shown by simulation results.
出处
《电子工程师》
2008年第10期31-34,共4页
Electronic Engineer
关键词
可修复系统
维修顺序
类状态
马尔可夫模型
映射算法
repairable system
maintenance sequence
class state
Markov model
mapping algorithm
