某飞机系统故障率敏感性分析研究

目的 解决某飞机系统针对不同服役阶段和使用环境的故障率敏感性分析问题，发现影响系统故障率的重要因素。方法 结合Sobol方法和Kriging模型，发展一种适用于某飞机系统的故障率敏感性分析方法。该方法通过收集该飞机系统在不同使用年限和使用环境下的故障率数据，并将其作为训练样本训练Kriging模型成为满足系统故障率预测要求的黑箱函数，然后利用训练好的Kriging模型预测敏感性分析所需的样本故障率。结果 利用该方法对某飞机子系统进行了故障率敏感性分析，给出了影响系统故障率的各变量的主效应及全效应指标，发现平均温度和平均湿度为影响该系统故障率的重要因素。结论 该方法可以更加高效地计算故障率影响变量的敏感性指标，并给出相对重要度排序，可为飞机系统的故障预防、故障诊断和故障维修提供依据。

Sensitivity Analysis for Failure Rate of Certain Aircraft System

The work aims to solve the problem of failure rate sensitivity analysis of certain aircraft system in different service stages and service environments and find out the important factors affecting the failure rate. A failure rate sensitivity analysis method suitable for certain aircraft system was proposed by combining Sobol method and Kriging model. The failure rate data of the aircraft system in different service life and service environments were collected, and used as the samples to train the Kriging model as a black box function to meet the prediction requirements of system failure rate, and then the trained Kriging model was used to predict the sample failure rate required for sensitivity analysis. By this method, the failure rate sensitivity analysis of certain aircraft subsystem was carried out, and the main effect and total effect indexes of variables affecting the system failure rate were given. It was found that the average temperature and average humidity were the most important factors affecting the system failure rate. This method can calculate the sensitivity index of affecting variables of failure rate more efficiently, and give the order of relative importance, which can provide basis for failure prevention, failure diagnosis and failure maintenance of aircraft system.