小波降噪与BSS在航空发动机故障诊断中的应用  

Application of Wavelet Denoising and Blind Source Separation in Aero-engine Fault Diagnosis

作　　者：马建仓[1] 陈静[1] 刘小龙[1] 

机构地区：[1]西北工业大学电子信息学院,陕西西安710129

出　　处：《计算机测量与控制》2009年第11期2115-2117,共3页Computer Measurement &Control

基　　金：国家自然科学基金项目(60672184)

摘　　要：在航空发动机的故障诊断中,传感器测得的信号通常是非平稳的振动信号;受发动机工作环境影响,这些振动信号含有大量噪声且多路源信号相互混叠;传统的信号处理方法很难从此类信号中快速有效地提取出故障特征;运用小波阈值降噪结合盲源分离的方法对发动机振动信号进行了分析,并对某型航空涡扇发动机发生空中停车故障时的振动信号进行了分析,验证了该方法在航空发动机故障诊断中的有效性。In aero--engine fault diagnosis, the signals from sensors are non--stationary vibration signals. The signals have numerous noises and the source signals mixed with each other. The traditional signal processing methods can＇t extract fault features fast and effectively. This paper used a method which combines blind source separation with wavelet threshold value denoising to analyze vibration signals, and processed vibration signals of a certain turbofan that appeared an in--flight shut--down failure to prove the validity of this method in aero-- engine fault diagnosis.

关 键 词：航空发动机 盲源分离 小波降噪 特征提取 故障诊断 

分 类 号：TN911.7[电子电信—通信与信息系统]