一种结合Bayes估计的NHPP航班查询软件可靠性评估方法

从航班查询系统故障的监控日志出发,分析了航班查询系统的故障数与其发生的时间间隔区间的关系,建立了NHPP模型。为了使NHPP模型能够更加准确地反映时间对航班查询系统的故障数的影响,对航班查询系统的寿命总体分布函数F(t)进行贝叶斯估计。当不重复采样时,贝叶斯估计取F(t)的先验分布为DP(α,P0),使得F(t)后验分布容易分析计算,从而使NHPP模型的估计更为精确。利用实际监控数据进行仿真,对模型进行了实现,结果表明了该模型对可靠性评估有较高的准确度。

A NHPP model for the open availability system based on Bayes estimation

From the fault monitoring logs of the Open Availability system,the relationship between the number of its faults and the time interval is analyzed, and a NHPP model is established based on Bayes estimation. Based on prior knowledge, Bayes estimation method deduces the life distribution function of the Open Availability system. When the sampling is not done repeatedly, Bayes estimation makes the prior distribution of F(t) as DP(α,P0),so the posterior distribution of F(t) is easily computed and thus the NHPP model which could accurately reflect the influence of the time interval on the number of faults of the Open Availability system, is improve. Real monitoring logs are used in the experiment to evaluate the model. Simulation results validate the correctness and effectiveness of this model.