摘要
针对航空发动机转静碰摩故障诊断问题,建立了含不平衡-碰摩故障的转子-滚动轴承耦合系统动力学模型。在模型中,充分考虑了滚动轴承的间隙、非线性赫兹接触以及变柔性VC(Varying compliance)振动。首先运用数值积分获取系统响应,进行碰摩故障分析,并得到了大量的碰摩故障仿真样本;其次利用支持向量机从大量样本中获取碰摩故障知识;然后利用转子实验器获取碰摩故障实验样本;最后利用训练好的支持向量机对碰摩故障实验样本进行智能诊断,最高识别率达到了91%。
Here,in order to diagnose rotor-to station element impact-rub fault in an aero-engine,a dynamic model of an unbalanced and rubbing rotor supported on ball bearings is established.In the model,clearance of bearing,nonlinear Hertzian contact force between balls and races and varying compliance vibration are considered fully.Firstly,the numerical integration method is employed to obtain the system's responses and many simulation samples of rubbing fault are obtained.Secondly,the simulation samples are employed to train a SVM.Then,the experimental samples can be obtained using the aero-engine tester.Finally,the intelligent diagnosis of rubbing fault using the trained SVM is performed and the highest recognition rate reaches 91%.
出处
《振动与冲击》
EI
CSCD
北大核心
2008年第10期90-94,114,共6页
Journal of Vibration and Shock
基金
国家自然科学基金资助项目(50705042)
航空科学基金资助项目(2007ZB52022)