Fault Diagnosis of Rolling Bearing Based on Relevance Vector Machine and Kernel Principal Component Analysis
In order to improve the speed and accuracy of rolling bearing fault diagnosis on small samples,a method based on relevant vector machine (RVM) and Kernel Principle Component Analysis (KPCA) is proposed.Firstly,wavelet packet energy of rolling bearing vibration signal is extracted with the wavelet packet transform which is used as fault feature vectors.Secondly,the dimension of feature vectors is reduced in order to educe the correlation between the features.The important principal components are selected by KPCA as new primary feature vectors,which the cumulative variance is greater than 95% as the criterion.Finally,combined with the relevant support vector machine,the fault of rolling bearing is diagnosed.The experimental results show that wavelet packet energy can express the rolling bearing fault features accurately,KPCA can reduce the dimension of feature vectors effectively,and the proposed method has better performance in the speed and accuracy of fault diagnosis than the one based on support vector machine (SVM),which supplies a strategy of fault diagnosis for rolling bearing.
rolling bearing relevant vector machine kernel principle component analysis fault diagnosis
Bo Wang Shulin Liu Hongli Zhang Chao Jiang
College of Mechatronics Engineering and Automation Shanghai University Shanghai,China;College of Mec College of Mechatronics Engineering and Automation Shanghai University Shanghai,China
国际会议
南京
英文
1-5
2013-08-20（万方平台首次上网日期，不代表论文的发表时间）