基于VMD包络相关系数的齿轮故障特征提取方法  

Fault Features Extraction Method of Gear Based on VMD Envelope Correlation Coefficient

作　　者：张超[1] 任杰[1] 何闯进 Zhang Chao;Ren Jie;He Chuangjin(North China Electricity Power University(Baoding),Baoding 071003,China)

机构地区：[1]华北电力大学(保定),河北保定071003

出　　处：《煤矿机械》2020年第10期168-171,共4页Coal Mine Machinery

基　　金：国家自然科学基金资助项目(51675178);中央高校基本科研业务费专项资金资助项目(2017MS152)。

摘　　要：使用变分模态分解(VMD)方法分析风力发电机齿轮故障信号时,本征模态函数(IMF)的个数K和约束参数α的选取对结果的影响很大,如何筛选合适的分解参数并获得最优分量是制约VMD实际应用的一个难题。提出一种基于包络相关系数的VMD参数筛选方法,能够针对风力发电机齿轮故障信号自动筛选出最佳VMD参数组合,得到最佳的IMF分量。与常用的齿轮故障诊断方法相比,该方法能够获得更好的分解效果,强化了故障信息,故障特征提取效果更好,为VMD方法的应用提供了新的思路。In the process of fault signal analysis for wind turbine gears using variational mode decomposition(VMD)method,the numbers K of the intrinsic mode functions(IMF)and the constrained parameterαboth have great influences on the results.It’s a difficult problem to select appropriate parameters of VMD to obtain the optimal components,which restricts the practical application of VMD.Proposed a method of VMD parameter selection based on an improved algorithm of envelope correlation coefficient which can automatically select the best combination of VMD parameters for gear fault signals of wind turbines to get the best IMF component.Compared with the commonly used methods of gear fault diagnosis,this method can obtain better decomposition effect,strengthen the information of fault and the extraction effect of fault features is better,which provides a new idea for the application of VMD method.

关 键 词：VMD 相关系数 齿轮故障 

分 类 号：TP133[自动化与计算机技术—控制理论与控制工程]