基于变分模态分解与双谱估计的滑动轴承摩擦状态评估  

Evaluation of Friction State of Plain Bearing Based on Variational Mode Decomposition and Bispectrum Estimation

作　　者：姜乃铭 Jiang Naiming

机构地区：[1]宁波大学机械工程与力学学院,浙江宁波315211

出　　处：《机械制造》2020年第6期1-4,共4页Machinery

基　　金：浙江省自然科学基金资助项目(编号:LY12E09001);宁波市自然科学基金资助项目(编号:2015A610150)。

摘　　要：针对滑动轴承工作中常见的摩擦失效问题,提出一种基于变分模态分解与双谱估计的电机电流特征分析方法,用于对滑动轴承的摩擦状态进行评估。应用这一方法,采集电机定子电流信号,对变分模态分解后的中高频信号进行重构,实现基频等无用信号的分离,提取摩擦特征量。摩擦特征多表现为非线性特征,双谱估计对非线性特征非常敏感,因此对重构后的中高频信号进行双谱估计。建立滑动轴承试验装置,通过试验表明随着摩擦的延续,双谱估计三维图发生明显变化,由此验证基于变分模态分解与双谱估计对滑动轴承摩擦状态进行评估的有效性。Aiming at the common problem of friction failure in the operation of the plain bearing,an analysis approach for motor current characters based on variational mode decomposition and bispectrum estimation was proposed to evaluate the friction state of the plain bearing.This method is applied to collect the current signal of the motor stator,and reconstruct the mid and high frequency signals after variational modal decomposition to realize the separation of garbage signals such as the fundamental frequency and extract the friction characteristic quantity.In most cases the friction features non linear behavior and bispectrum estimation is very sensitive to non linear features.Therefore,bispectrum estimation is performed on the reconstructed mid and high frequency signals.A test device for the plain bearing was established,and experiments showed that with the continuation of friction,the three dimensional map of bispectrum estimation changed significantly,thereby the effectiveness of the evaluation for friction state of the plain bearing based on variational mode decomposition and bispectrum estimation was verified.

关 键 词：滑动轴承 变分模态分解 双谱估计 摩擦 

分 类 号：TH133.31[机械工程—机械制造及自动化]