中文摘要：

结合物理模型与灰色理论,提出行星轮系齿根疲劳裂纹故障预测的新思路。针对直升机主传动系统中的2K-H行星轮系,建立太阳轮齿根疲劳裂纹损伤的物理基模型,通过仿真获得不同损伤严重度的振动仿真信号。选择并计算仿真信号的故障特征矢量,并以此作为损伤特征的标准模式,对待检信号特征矢量与标准模式进行灰色关联度分析,根据关联度对裂纹进行定量检测。结合物理模型仿真信号对灰色预测模型进行修正,使之具有更好的疲劳裂纹故障预测能力。对试验中的疲劳裂纹进行定量检测和故障预测。试验数据验证了本方法对行星轮系太阳轮疲劳裂纹的定量检测和故障预测能力。

英文摘要：

A new idea of predicting fatigue crack in planetary gear sets is presented,which is based on physical model and gray theory.Aiming at 2K-H planetary gear sets,physics-based models for damaged system with multi-stage severity are established,and simulation data of damaged system is obtained.Feature vector of simulation data selected and computed is used as standard damage mode.Grey correlation analysis is carried out between standard damage mode and the feature vector of signal to be checked.Quantitative detection of cracks is carried out on the basis of the correlation degree.By the aid of simulation data,the grey model is modified so as to have a better performance in predicting fatigue crack.Seeded fault tests are carried out to validate the methodology and a promising result is acquired.