摘要
针对传统的时域、频域特征不能明显地表征滚动轴承的早期退化特征的问题,提出了一种小波包能量谱结合主成分分析构建综合评估函数的滚动轴承早期性能退化评估方法。该方法以采集到的轴承正常工作时的振动信号作为训练样本,对样本进行小波包能量谱计算,得到高维特征向量;再利用主成分分析方法降维并建立综合评估函数对早期性能退化区的数据进行判断。运用实测的滚动轴承全寿命实验数据进行检验,结果表明该方法能实现对滚动轴承早期性能退化的评估。
The traditional time-frequency characteristics was not obvious in the early degradation extraction of the rolling bearings so this paper put forward a method based on wavelet packet energy spectrum in combination with principal component analysis for the early degradation extraction.First using normal signal data the wavelet packet energy spectrum was acquireal,then the principal component analysis was used to achieve dimension reduction for the high dimensional feature vector.Finally through the principal component analysis the comprehensive evaluation function was constructed,which was used to achieve the early degradation extraction of the rolling bearings.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2015年第17期2352-2355,2368,共5页
China Mechanical Engineering
基金
国家自然科学基金资助项目(51275546
51375514)
高等学校博士学科点专项基金资助项目(20130191130001)
关键词
小波包能量谱
主成分分析
早期退化评估
综合评估函数
wavelet packet energy spectrum
principal component analysis
early degradation assessment
comprehensive evaluation function
