基于BB算法的左截断威布尔分布可靠性分析

针对寿命分布类型和参数估计方法问题,对国内某航空公司现役机型B737-800飞机主起落架减振支柱漏油故障数据进行可靠性分析。结合数据类型特点,初步预判其寿命分布类型可能为左截断两参数威布尔分布、对数正态分布、指数分布。基于极大似然估计法分别估计三种分布的未知参数,并利用根据BB算法改进的柯尔莫哥洛夫-斯米尔诺夫假设检验(K-S检验)法判断其寿命分布类型。实证分析结果表明,采用左截断两参数威布尔分布来拟合是合适的,并提出了具体维修建议。

Reliability Analysis of Left-truncated Weibull Distribution Modified by BB Algorithm

A reliability analysis was done based on the failure data of B737-800 main landing gear shock strut oil leakages herein, which came from a domestic airlines' electronic records, in consideration of the life distribution type and parameter estimation method. The characteristics of data preliminarily demonstrated that the life distribution type might be left-truncated two parameter Weibull distribution, logarithmic normal distribution or exponential distribution. Next, the parameters corresponding to three types of distributions were estimated by using the MLE method. Then, the Kolmogorov-Smirnov(K-S) hypothesis test, which modified by BB algorithm and goodness-of-fit test , was used to identify and confirm the life distribution function model. The analysis results show that the distribution of the data is left-truncated two parameter Weibull distribution. In the end, reasonable maintenance plan and optimized maintenance scheme are provided.