基于HHT的航空发动机转子故障特征自动提取与智能诊断  

Fault Characteristics Automatic Extraction and Intelligent Diagnosis of Aero-engine Rotor Based on HHT

作　　者：高斌[1] 

机构地区：[1]滨州学院飞行学院,山东滨州256603

出　　处：《滨州学院学报》2015年第6期19-22,共4页Journal of Binzhou University

基　　金：滨州学院科研基金项目(BZXYG1208)

摘　　要：提出了一种基于希尔伯特—黄变换的航空发动机转子故障特征自动提取方法,该方法利用经验模态分解得到转子故障信号的各模态分量,结合故障频率特征,直接从各模态分量中提取反映转子故障特征的特征向量。建立基于结构自适应神经网络方法的转子故障智能诊断模型,利用实验数据进行了实验验证,结果表明本方法在转子故障智能诊断方面取得了较好的效果。This paper presents an automatic extraction method for aero-engine rotor fault characteris- tics based on HHT. IMFs of rotor fault signal are got with EMD. Combined with fault frequency charac- teristics, the eigenvector which reflected rotor fault characteristics is extracted from IMFs. The integrat- ed neural network intelligent model for rotor fault based on the neural networks model with self-adaptive structure is established. The practical aero-engine rotor experiment data is used to verify the new method which is put forward in this paper, and the results fully show its correctness in terms of rotor fault intel- ligent diagnosis.

关 键 词：希尔伯特—黄变换 转子故障 神经网络 频率特征 智能诊断 

分 类 号：V231[航空宇航科学与技术—航空宇航推进理论与工程]