基于Weibull混合分布的航空发动机零部件寿命建模

Aeroengine Component Life Modeling Based on Mixed Weibull Distribution

针对航空发动机在维修工作中存在重要零部件使用寿命及失效率难以估算的问题,搜集了发动机重要零部件燃油计量活门、整体驱动发电机、液压机械组件的使用寿命数据,并运用Weibull混合分布及单Weibull分布分别建立了上述重要零部件的寿命模型。在模型求解过程中采用了经典的三参数相关系数优化法,求出发动机重要零部件的可靠度、失效率及故障概率密度函数。通过计算得出Weibull混合分布模型的线性相关系数r_(g)比单Weibull分布模型的大;同时K-S假设检验结果显示,Weibull混合分布的统计量观测值D_(n)均小于临界值D_(30,0.1)。结果表明:相对于单Weibull分布,Weibull混合分布模型更精确地表征了发动机重要零部件的寿命数据,从而有效地提高寿命参数的估算精度;Weibull混合分布模型的可靠性有助于发动机管理工程师制定视情维修方案;并对航材备件的数量有指导意义,具有较高的推广应用价值。

In response to the difficulties in estimating the service life and failure rate of important components during aeroengine maintenance,the service life data of important aeroengine components were collected,including the fuel metering valve,the integral drive generator,and hydraulic mechanical components.Life models of these components were established by using mixed Weibull distribution and single Weibull distribution respectively.In the process of solving the model,the classical three-parameter correlation coefficient optimization method was used to find the reliability,failure rate,and failure probability density function of these components.The results show that the linear correlation coefficient r_(g) of the mixed Weibull distribution model is larger than that of the single Weibull distribution model.Meanwhile,the K-S hypothesis test results show that the observed statistical values D_(n) of the mixed Weibull distributions are all less than the critical value D_(30,0.1).The above results show that compared with the single Weibull distribution,the mixed Weibull distribu⁃tion model can more accurately characterize the life data of important engine components,thus effectively improving the estimation accuracy of life parameters.The mixed Weibull distribution model,owing to its reliability,is helpful for engine management engineers to make condition-based maintenance plans and can be used in guiding the determination of the number of spare parts which is worthy of extensive application.