基于EWT-SVDP的旋转机械故障诊断  

Rotating machinery fault diagnosis based on improved atomic sparse decomposition algorithm

作　　者：王娜[1] WANG Na(Xi‘an Aeronautical Polytechnic Institute,Xi’an 710089,CHN)

机构地区：[1]西安航空职业技术学院,陕西西安710089

出　　处：《制造技术与机床》2018年第11期73-78,共6页Manufacturing Technology & Machine Tool

基　　金：国家自然科学基金资助项目(51105323)

摘　　要：为了准确诊断旋转机械故障类型,提出了基于EWT-SVDP的故障诊断方法。分析了经验小波变换对信号分解的频率自适应性优势;给出了奇异值包分解与重构原理,并说明了此算法的强去噪能力;将经验小波变换与奇异值包分解融合,提出了EWT-SVDP算法,此算法兼容了经验小波变换的频率自适应性与奇异值包分解的强去噪能力,能够有效去除信号误差并给出信号频谱;介绍了轴承基本结构和故障特征频率理论值,设计了轴承故障诊断实验方案和方法,对故障振动信号使用EWT-SVDP算法进行去噪和重构,并结合希尔伯特变换,分析信号的频谱和包络图,可以明显看出轴承回转频率及倍频、故障特征频率及其倍频、两者合成的边频等,充分证明了所提出的EWTSVDP算法能够准确判断出旋转机械故障类型。To diagnose type of rotating machinery fault accurately,fault diagnosis method based on EWT-SVDP is proposed.Frequency adaptation decompose of empirical wavelet transform are analyzed.Strong denoising capability and detail analysis capability of Singular Wavelet Packet Decomposition are given.Combined EWT with SVDP,EWT-SVDP is put forward which possess both advantages.Bearing basic structure and fault character frequency are introduced.Bearing fault diagnosis experiment scheme is designed.By EWT-SVDP decomposition and reconstitution of fault vibration signal,bearing slew frequency and its multiplication,fault character frequency and its multiplication,and side frequency of the both are obvious,which can prove that EWT-SVDP can figure out rotating machinery fault obviously.

关 键 词：旋转机械 故障诊断 经验小波变换 频率自适应 奇异值包分解 

分 类 号：TP17[自动化与计算机技术—控制理论与控制工程]