论文部分内容阅读
设备冷却水泵具有高可靠性、长寿命的特点,历史失效数据的小子样问题突出,并且由冲击、振动、磨损、腐蚀等耗损性因素引起的故障,其故障率具有时变性,随时间的推移而增加.故障率λ为常数的Jeffreys先验模型不能以合理概率复现设备冷却水泵的故障数据,无法满足分析λ时变性的要求.本文在对数线性模型基础上为λ引入额外时变性,建立Poisson分布的广义线性模型,研究λ的时间趋势;不仅利用定性图检验评价模型复现设备冷却水泵故障数据的能力,而且还应用贝叶斯χ~2统计量进行定量化检验;经过定性和定量的双重检验,表明该模型据有良好的预计能力,可以分析设备冷却水泵故障率λ的时变性.
Equipment cooling water pump has the characteristics of high reliability and long life. The problem of small sample of historical failure data is prominent, and the failure rate caused by the wear and tear factors such as impact, vibration, abrasion and corrosion has a time-varying failure rate. With the passage of time, While the failure rate λ is constant, the Jeffreys a priori model can not reproduce the fault data of the cooling pump of the equipment with a reasonable probability and can not meet the requirement of analyzing the time-dependent variation of lambda.In this paper, extra time-varying is introduced for λ on the basis of the logarithm linear model, The generalized linear model of Poisson distribution is established and the time trend of λ is studied. The qualitative analysis of the model is not only used to evaluate the ability of the device to reproduce the fault data of cooling water pump, but also to be quantified by Bayesian χ ~ 2 statistic. Quantitative double test shows that the model has good predictive ability and can analyze the time-varying of failure rate λ of equipment cooling water pump.