基于奇异值分解和支持向量机的转子故障诊断  

Rotor Fault Diagnosis Based on Singular Value Decomposition and Support Vector Machine

作　　者：柴凯[1] 张梅军[1] 黄杰[1] 陈灏[1] 

机构地区：[1]解放军理工大学野战工程学院,江苏南京210007

出　　处：《机械工程与自动化》2014年第4期120-122,共3页Mechanical Engineering & Automation

基　　金：国家自然科学基金资助项目(51175511)

摘　　要：为了准确识别转子不平衡、不对中、碰摩和油膜涡动等故障,利用小波分析对转子故障信号进行4层分解,将频率由高到低的5个分支信号作为奇异值分解(Singular Value Decomposition,SVD)矩阵的行向量,经奇异值分解后得到信号的故障特征值。通过支持向量机(Support Vector Machine,SVM)在选择不同的核函数和结构参数下比较其对转子故障诊断结果的影响。结果表明在选择最优SVM模型和参数的基础上,对SVD获得的故障特征值进行诊断,得出了准确的诊断结果。In order to accurately identify the rotor faults of imbalance,misalignment,rubbing and oil whirl,this paper adopts wavelet analysis method to decompose the fault signal to obtain the five branches from high to low frequency as the row vectors of SVD（Singular Value Decomposition）matrix,the characteristic value of fault signal can be obtained after singular value decomposition.By using SVM（Support Vector Machine）to select different kernel functions and structural parameters,the paper studies the influence of different kernel functions and structural parameters on rotor fault diagnosis result.The results show that selecting optimal SVM model and parameters to analyze the characteristic value obtained by SVD can get an accurate diagnosis result.

关 键 词：奇异值分解 支持向量机 故障诊断 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]