基于敏感分量融合的液压泵退化特征提取方法  

A degradation feature extraction method for hydraulic pumps based on fusion of sensitive components

作　　者：孙健[1] 李洪儒[1] 田再克[1] 

机构地区：[1]军械工程学院,石家庄050003

出　　处：《仪器仪表学报》2016年第6期1290-1298,共9页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(51275524)项目资助

摘　　要：液压泵退化特征提取是实现故障预测的关键环节。在液压泵性能退化过程中,其振动信号复杂度高、非线性强,难以有效地提取退化特征,为此,本文提出一种基于敏感分量融合的退化特征提取方法。采用改进局部特征尺度分解(ILCD)方法对振动信号进行分解,并利用贝叶斯信息准则与所构建的敏感因子,对内禀尺度分量进行筛选,以减少干扰分量的影响,得到敏感分量,有效抓取特征信息;在此基础上,引入离散余弦变换代替传统复合谱分析中的傅里叶变换,提出离散余弦变换-复合谱(DCS)算法,以解决信息遗漏问题,并利用DCS对敏感分量进行融合,提取复合谱熵作为退化特征,以提高对退化过程的表征能力;最后,通过对液压泵性能退化试验实测振动信号的应用分析,验证了该方法的有效性。Degradation feature extraction of hydraulic pump is the key step for prognostics. During hydraulic pump degradation, the vibration signal is of high complex and nonlinear, and it is difficult to extract effective degradation features. Therefore, a novel degradation feature extraction method based on the fusion of sensitive components is proposed. Vibration signal is firstly decomposed by Improved Local Characteristic-scale Decomposition （ILCD）. And the Intrinsic scale components （ISC） are selected based on Bayesian Information Criterion and the presented sensitive factor, to reduce unnecessary components and obtain feature information effectively. In further, the Fourier Transformation is substituted with Discrete Cosine Transform （DCT） in traditional composite spectrum. Thus, the DCT-Composite Spectrum （DCS） algorithm is proposed so as to reduce information missing and improve feature sensitivity. The achieved sensitive components are further fused by DCS, and the composite spectrum entropy is extracted as degradation feature. Finally, the proposed method is verified by vibration signal in the hydraulic pump degradation experiments.

关 键 词：退化特征提取 ILCD 复合谱算法 液压泵 

分 类 号：TN911.7[电子电信—通信与信息系统] TH137.51[电子电信—信息与通信工程]