基于多特征融合与XGBoost的风机轴承故障诊断

Fault diagnosis of wind turbine bearing based on multi-feature fusion and XGBoost

针对单独从时域、频域、时频域对风机轴承振动信息描述的不充分性,以及传统故障诊断算法精度较低的问题,提出了基于多特征融合与XGBoost的风机轴承故障诊断算法。首先,提取轴承振动信号的时域、频域以及时频域特征,然后利用XGBoost算法对轴承故障进行诊断,计算每个特征在树节点分裂中获取的信息增益,并根据特征信息增益对特征进行筛选,最后采用支持向量机(SVM),K最近邻(KNN),人工神经网络(ANN)等算法对筛选后的特征进行故障诊断对比实验。仿真结果表明:本文算法可以提取出具有高区分性和独立性的特征,同时在轴承故障诊断率上优于其他算法。

A wind turbine bearing fault diagnosis method based on multi-feature integration and extreme gradient boosting(XGBoost)is proposed in addressing the insufficient describing bearing vibration information from time domain,frequency domain and time-frequency domain separately,and the low accuracy in traditional fault diagnosis algorithms.Firstly,the time domain,frequency domain and time-frequency domain characteristics of the bearing vibration signal are extracted.Then XGBoost algorithms is used to diagnose the bearing faults.The information gain of each feature in tree node splitting is calculated,and features are selected according to the information gain.Finally,using support vector machine(SVM),K-nearest neighbor(KNN),artificial neural network(ANN)and other algorithms to conduct fault diagnosis and comparison experiments on the selected features.The simulation results show that the proposed fault diagnosis algorithm can extract features with high distinction and independence,and is superior to SVM,KNN,ANN and other algorithms in bearing fault diagnosis rate.