带有线性退化和随机振荡的系统可靠性及维修决策  

Reliability and Maintenance Decision for Systems Subjected to Linear Degradation and Random Shocks

作　　者：张斌[1] 费文龙[1] 张岚[1] 

机构地区：[1]南京信息工程大学数学与统计学院,江苏南京210044

出　　处：《数理统计与管理》2016年第4期669-678,共10页Journal of Applied Statistics and Management

摘　　要：研究了随时间发生线性退化和随机振荡导致瞬时退化的系统可靠度及定期维修策略。随机振荡的发生次数服从非时齐泊松过程,每次振荡造成系统的退化量独立同分布。当累积退化量达到阀值时,系统发生故障。为了改善系统工作状态,降低故障风险,每隔T时对系统进行不完全预防维修,维修后故障率函数将发生变化,维修成本与系统的退化程度有关。在NT时,对系统进行完全预防维修,使系统修旧如新。构建了系统的可靠度函数。在单位时间平均利润最大的前提下,提出不完全预防维修间隔T和完全预防维修周期NT的确定方法。分析了模型参数对维修决策的影响。This paper studies the policies of the periodic maintenance for a system subjected to bothlinear degradation and random shocks. Random shocks arrive according to a non-homogeneous Poissonprocess. The damage sizes caused by the shock are i.i.d, random variables. A failure occurs whenthe overall degradation is beyond a threshold value. We consider two different preventive maintenancepolicies： Imperfect preventive maintenance （IPM） that is performed at periodic intervals and perfectpreventive maintenance （PPM） that fully upgrades the system to the as-good-as-new condition. Thehazard-rate function after IPM will be changed. The cost of IPM depends on the total degradation of thesystem. A generM reliability model is developed. The optimal IPM interval and PPM cycle that maximizethe expected profit per unit of time are derived. The effects of process parameters on maintenance policiesare studied.

关 键 词：线性退化 随机振荡 可靠度 维修决策 

分 类 号：O212[理学—概率论与数理统计]