无失效数据下航空轴承的可靠性分析

针对某型航空轴承可靠性试验过程中出现的无失效问题,在对轴承寿命分布进行分析的基础上,首先根据多层Bayes法,研究了无失效可靠性试验中的试验分组问题,发现分组数在6~9之间选择时评估效果相对较好,且航空轴承的寿命越长分组数可以越少.其次,分析了多层Bayes法和E-Bayes法的原理和计算过程,通过Monte Carlo仿真,以最佳分组数为基础,发现两种方法的估计结果相差不大,但从可靠度角度综合来看,E-Bayes法相对较优.最后讨论了超先验分布中参数c的选值问题,通过对比分析给出了相对较适合于航空轴承无失效数据分析时的参数值.

Reliability analysis for aircraft bearing with zero-failure data

For the zero-failure conditions of the aircraft bearing that may occur during the process of reliability test, on the basis of analysis about distribution of the aircraft bearing life, the problem of the test groupings was firstly analyzed according to the hierarchical Bayesian method; it was found that when the number of groupings in 6-9 was chosen, the effects of evaluation were relatively better. And the longer life of aircraft bearing indicated less number of groupings. Second, based on the optimum groupings, by analyzing the hierarchical Bayesian method and E-Bayes method, it was proved that the estimation results of the two methods were similar, but from the point of comprehensive reliability, E-Bayes method was better by comparing the estimation results based on the zero-failure data from the simulation values by Monte Carlo. Finally, the selected values of parameter c of the super prior distribution were discussed. Through comparison analysis, the relatively appropriate parameter values suitable for the aircraft bearing were found, providing a theoretical guidance for the reliability analysis of zero-failure data. An enterprise case was given to verify the conclusion.