基于贝叶斯分析的不完全维修条件下可修系统的参数估计

Parameter Estimation for a Repairable System under Imperfect Maintenance Based on Bayesian Analysis

可靠性和维修性的参数估计在可修系统建模及维修决策中具有非常重要的地位.不完全维修条件下可修系统的故障时间既不相等也不服从独立分布,这使得可修系统维修模型的参数估计变得十分困难.针对此问题,对第一次故障服从威布尔分布,故障之后进行Kijima类型I的修复性维修和Kijima类型I的预防性维修的可修系统进行了研究,应用贝叶斯分析对威布尔分布的形状参数β、尺度参数η、修复性维修的不完全维修因子ar以及预防性维修的不完全维修因子ap进行估计.仿真算例表明,该方法可以较为准确的对可靠性和维修性参数进行估计.

Estimation of reliability and maintainability parameters is important in modeling of repairable systems and determining maintenance policies. The failure times of repairable systems under imperfect maintenance are neither identically nor independently distributed, which makes parameter estimation difficult. The paper research on the repairable system, which follows a Weibull distribution in the first failure and experiences Kijiama Type？imperfect corrective maintenance and Kijiama Type？imperfect preventive maintenance after failure, and estimates the shape parameter of the Weibull distribution β, the scale parameter of the Weibull distribution 71, the imperfect maintenance factor for preventive maintenance ar and the imperfect maintenance factor for preventive maintenance ap. The results of simulated example indicate the proposed method is precision to estimate the reliability and maintainability parameters.