中国民航大学学报

• 民用航空 • 上一篇    下一篇

基于小波神经网络的航空发动机故障诊断研究

杨永刚,顾杰   

  1. (中国民航大学中欧航空工程师学院,天津300300)
  • 收稿日期:2015-11-18 修回日期:2016-01-18 出版日期:2016-10-19 发布日期:2016-12-06
  • 作者简介:杨永刚(1978—),男,吉林梨树人,副教授,博士,研究方向为故障诊断、机器人技术.
  • 基金资助:

    国家自然科学基金项目(61172013)

Study on aero-engine faults diagnosis based on wavelet neural network

YANG Yonggang, GU Jie   

  1. (Sino-Europewn Institute of Aviation Engineering, CAUC, Tianjin 300300, China)
  • Received:2015-11-18 Revised:2016-01-18 Online:2016-10-19 Published:2016-12-06

摘要:

针对航空发动机的一些常见故障类型,为了达到进行有效识别的目的,提出了一种基于小波包和神经网络相结合的发动机故障诊断方法。以某型航空发动机为研究对象,通过小波包对采集到的振动数据进行分解和重构,提取出表征发动机工作状况的特征向量,并将其作为训练样本数据和检验样本数据,输入小波El-man 神经网络中并对其进行训练,试验结果表明:这种诊断模型能够有效地识别出所研究的航空发动机故障类型,故障诊断率较高。

关键词: 小波变换, 神经网络, 航空发动机, 故障诊断

Abstract:

In order to effectively identify the common fault types of aviation engine, a new method of engine fault diagnosis based on wavelet packet and neural network is proposed. Taking a certain type of aero-engine as research object,the vibration signal is decomposed and reconstructed through wavelet packet to obtain feature vector of its working condition. Then these data are input into the wavelet Elman neural network as training sample data and test sample data. Experimental results show that this method is feasible and the fault types of aero-engine is well recognized.

Key words: wavelet transform, neural network, aero-engine, fault diagnosis

中图分类号: