基于贝叶斯模型的装备剩余寿命预测研究  

Bayesian Model for Predicting Residual Life of Equipment

作　　者：刘刚[1,2] 黎放[1,2] 

机构地区：[1]海军工程大学,武汉430033 [2]海军工程大学舰船综合保障工程研究室,武汉430033

出　　处：《火力与指挥控制》2016年第5期19-24,共6页Fire Control & Command Control

摘　　要：为了解决武器装备日益复杂及维修工作日趋繁重的问题,运用贝叶斯模型预测装备修理后的剩余寿命,为合理安排其修理计划提供依据。替代传统的指数分布,用威布尔分布描述系统寿命特征,并运用极大似然方法和贝叶斯方法估计威布尔分布的两个未知参数,给出其置信区间。在此基础上,对先验样本和后验样本两种不同情况,分别运用贝叶斯模型预测装备修理后的剩余寿命,并给出实例。结果表明了该方法的有效性。To solve the problem of weapon equipment is the growing complexity and themaintenance work becoming heavy daily,the Bayesian model is applied to predicting residual life ofequipment after repair,the foundation was proposed for reasonably arranging repair plan. Thetraditional exponential distribution is replaced,then Weibull distribution is used for describing lifecharacteristic of system,furthermore,and the methods of Maximum likelihood and Bayesian areapplied to estimating the two parameters of Weibull distribution,and the credible intervals are given.On the basis,according to the two situations that prior sample and posterior sample,the residual lifeafter repair is predicted by Bayesian model respectively. The illustrative example is given in paper,and the proposed method is proved to be effective.

关 键 词：威布尔分布 贝叶斯 剩余寿命 预测 

分 类 号：TP391.73[自动化与计算机技术—计算机应用技术]