综合应用小波包支持向量机直升机齿轮故障识别及实验研究

摘要

A new approach（WT-SWM） of feature extraction and fault recognition is presented based on the wavelet transform（WT） and support vector machines（SVM） to improve the performance of gear fault recognition for the helicopter transmission.Firstly,the noise is removed from the original time-domain vibration signals using the WT technique and energy features are extracted from those preprocessed.Afterwards,the energy features are used as inputs to the classifier of SWM for two-class（normal or fault） recognition.A gear drive failure test rig is designed and constructed,with which normal and defective gears of spiral bevel gear pair vibrative test are processed.The data processing results show that the present classification accuracy of the WT-SWM method for the fault spiral bevel gear diagnosis is 100% that is higher than compared well with the results of the ANN only.%提出了一种综合运用小波包与支持向量机,对直升机主减速器齿轮进行特征提取与故障识别的方法。首先利用小波分析对齿轮振动信号阈值去噪,去除掉系统噪声的影响;然后在离散小波变换信号分解的基础上提取信号能量特征,利用支持向量机对齿轮系统的故障进行特征信息识别。建立了弧齿锥齿轮传动系统振动测试试验台,对正常结构和故障结构的传动系统分别进行了试验测试和振动信号的数据处理。结果表明：该方法对齿轮故障特征识别结果的有效率可达100%。