寿命指数分布类型装备可靠性指标估计方法的应用分析

Application Analysis of Reliability Index Estimation Method for Exponential Distribution Type Equipment

针对目前舰船电子元器件等寿命服从指数分布类型的装备在进行可靠性参数指标评价时,缺少试验数据样本量、参数估计方法较难确定以及理论参数值设置准确性差等问题,在寿命分布确定的情况下,基于装备使用数据,开展了运用极大似然估计和Bayes估计对指数分布类型装备可靠性参数进行估计的应用研究。分析了两种估计方法的计算准确性及先验信息的准确程度对计算结果的影响;比较了两种参数估计方法的估计结果,得到结论:在理论MTBF值和实际MTBF值接近的情况下,采用Bayes估计方法效果更理想,当理论MTBF值和实际MTBF值偏差较大时,采用极大似然估计方法更为合适;同时估算出了临界值,可用于推断装备设置的理论MTBF值是否准确,对于提高装备可靠性指标评价方法的可行性、准确性和适用性具有重要的指导意义。

Aiming at the problems of the current life-exponential distribution type equipment such as ship electronic components when evaluating reliability parameter indicators,the lack of test data sample sizes,parameter estimation methods are difficult to determine,and the theoretical parameter value settings are accurate,etc.,an application study was carried out to estimate the reliability parameters of the exponential distribution type equipment using maximum likelihood estimation and Bayes estimation when the life distribution is determined,based on the equipment usage data.The calculation accuracy of the two estimation methods and the influence of the accuracy of the prior information on the calculation results were analyzed;The results of the two parameter estimation methods were compared and the conclusions were obtained:when the theoretical MTBF value is close to the actual MTBF value,the Bayesian estimation method is more effective.When the theoretical MTBF value and the actual MTBF value have a large deviation,the maximum likelihood estimation method is more suitable;At the same time,a critical value was estimated,which can be used to infer whether the theoretical MTBF value of the equipment setting is accurate,and it has important guiding significance for the feasibility,accuracy and applicability of the equipment reliability index evaluation method.