基于耦合双谱微分性质的微弱故障信息提取及应用  

Feature Extraction of Weak Fault Signal and Application Based on Coupled Bispectrum Differential Property

作　　者：梅二召 范景峰 肖玉玲 张翼飞 MEI Erzhao;FAN Jingfeng;XIAO Yuling;ZHAGN Yifei(School of Mechanical and Electrical Engineering,Henan Technical Institute,Zhengzhou Henan 450042,China)

机构地区：[1]河南应用技术职业学院机电工程学院,河南郑州450042

出　　处：《机床与液压》2021年第16期198-203,共6页Machine Tool & Hydraulics

基　　金：河南省科技攻关项目(212102210623);河南应用技术职业学院校级科研项目(2021-KJ-37,2020-KJ-26);河南省教育科学“十三五”规划项目(2020YB0500)。

摘　　要：为了提取减压阀中微弱的故障信息,利用减压阀振动信号中包含的耦合性质进行微分双谱运算来确定其故障状态。根据建立的AR双谱模型对微分前后的复数三阶累积量进行对比分析,微分后的AR双谱中所增加的频率信息对故障诊断结果产生了一定影响,经过数次微分后发现增加微分次数可以提取到信号中微弱的耦合信号成分从而能够提取到更多的故障信息。实验结果表明:通过改变微分次数,可以区分故障信息不明显的振动信号,从而提高故障诊断的正确率。In order to extract the weak fault information of the pressure relief valve,the differential bispectrum operation was used to determine the fault state of the pressure relief valve by using the coupling property contained in the vibration signal.According to the AR bispectrum model,the complex third-order cumulants before and after differentiation were compared and analyzed.The frequency information added in the AR bispectrum after differentiation had a certain influence on the result of fault diagnosis.After several differentials,it is found that the weak coupling signal components in the signal can be extracted by increasing the number of differentials,so more fault information can be extracted.The experimental results show that by changing the number of differential,the vibration signals whose fault information is not obvious can be distinguished,thus the accuracy of fault diagnosis is improved.

关 键 词：三阶累积量 耦合 微分 AR双谱 故障诊断 

分 类 号：TH137.52[机械工程—机械制造及自动化]