基于机群仿真模型的平均秩次法有效性分析

Analysis of validity with mean rank order method based on air fleet simulation model

为了验证平均秩次法在估计完全删失数据分布函数参数时的有效性,设计了一种明晰平均秩次法适用范围的蒙特卡罗仿真方法。该方法考虑现场故障数据的实际生成过程,以机群为研究对象,通过飞机架数、飞行天数、日利用率动态变化驱动故障样本、删失比变化,获取故障间隔时间的完全删失数据。结合拉依达准则,确定了平均秩次法的计算错误率。仿真结果表明,样本量在9.1,删失比为0.328时,错误率达到28.9%,此条件下,平均秩次法已经不适用。因此,在航空工程实际的可靠性评估中,由于大量高可靠性产品的出现,平均秩次法的适用范围变窄。

In order to validate the validity of the mean rank order method when estimating distribution {unction^s parameters of complete and censored data, a Monte Carlo simulation method to clear the applicable range of mean rank order method is designed. Given the actual process of recorded fault data, the complete and censored data of time between failures through the changes of numbers of aircraft, censored rates, days of flight and utili- zation rate of every day are gotten by using air fleet. The miscalculation probability of the mean time order method with Pauta criterion is calculated. The result shows that miscalculation of the mean rank order method is 28. 9% when fault samples are 9.1 and censored rates are 0. 328, in which the method is inadaptability. So application of the method becomes narrow in aviation engineering reliability assessment as the emergence of abundanthigh reliability product.