基于稀疏表示和SVM的航空发动机故障诊断  

Fault Diagnosis of Aircraft Engines Based on Sparse Representation and SVM

作　　者：巩孟林 陈卫[1] 钟也磐 杜炜[1] 李思路[1] 梁涛[1] GONG Meng-lin;CHEN Wei;ZHONG Ye-pan;DU Wei;LI Si-lu;LIANG Tao(Air Force Engineering University, Xi’an, 710038, China)

机构地区：[1]空军工程大学,西安710038

出　　处：《噪声与振动控制》2017年第3期162-167,共6页Noise and Vibration Control

基　　金：国家自然科学基金资助项目(51175509)

摘　　要：针对某型航空发动机减速器一级齿轮毂断裂问题,考虑其不易拆卸的特点,提出基于信号稀疏表示和支持向量机(Support Vector Machine,SVM)的故障诊断算法。首先,利用稀疏表示提取出最大和次大的稀疏系数作为特征向量。其次,选取支持向量机进行故障识别,在小样本学习条件下保持了较高的识别准确率。通过对简易减速器和航空发动机振动信号的分析证明了所提算法的有效性及其在工程应用中的价值。Considering the dismantling difficulty of the reducer of an aircraft engine and the necessity of the crack detection in its first grade gear hub,a fault diagnosis method based on sparse representation and support vector machine(SVM)is proposed.Firstly,the sparse representation is used to extract the largest and the secondary largest sparse factors as the feature vectors.Then,the fault is recognized using SVM,which maintains the high recognition accuracy under small training sample capacity condition.The analysis of vibration signals from a simple reducer and an aero-engine proves the efficiency and engineering application value of the proposed method.

关 键 词：振动与波 航空发动机 故障诊断 稀疏表示 支持向量机 

分 类 号：O422.6[理学—声学]