基于频率诱导变分模态分解的齿轮箱故障诊断  

Fault diagnosis of gearbox based on frequency-induced variational mode decomposition

作　　者：马天霆 孙振波 邓艾东[1,2] 邓敏强 Ma Tianting;Sun Zhenbo;Deng Aidong;Deng Minqiang(School of Energy and Environment,Southeast University,Nanjing 210096,China;National Engineering Research Center of Turbo-Generator Vibration,Southeast University,Nanjing 210096,China;Guoneng Taicang Power Plant Co.,Ltd.,Suzhou 215000,China)

机构地区：[1]东南大学能源与环境学院,南京210096 [2]东南大学火电机组振动国家工程研究中心,南京210096 [3]国能太仓发电有限公司,苏州215000

出　　处：《东南大学学报（自然科学版）》2023年第4期702-708,共7页Journal of Southeast University：Natural Science Edition

基　　金：国家自然科学基金资助项目(51875100,52005267).

摘　　要：针对风电传动系统齿轮箱的故障诊断问题,在脉冲激励响应的基础上提出了一种频率诱导变分模态分解(FIVMD)方法,并将其应用于齿轮箱故障特征提取.首先,根据振动信号傅里叶谱的极大值分布估计齿轮箱系统的自振频率;然后,将固有频率的估计值作为各模态分量中心频率的初始化位置,并通过交替乘子法将原始信号自适应分解为本征模态函数;其次,通过希尔伯特变换对各本征模态函数求包络谱,并计算其故障特征频率比;最后,挑选出故障特征频率比最大的模态分量,并根据其包络谱特征实现齿轮箱故障的有效识别.以维斯塔斯某3 MW风电机组圆柱齿轮断齿故障为例,验证了FIVMD在工程应用中的有效性和优越性.Aiming at the problem of fault diagnosis of the gearbox in wind turbine transmission chain,a frequency-induced variational mode decomposition(FIVMD)method is proposed based on the impulse excitation response and applied to the extraction of fault characteristics.First,the natural frequencies of the gearbox system are estimated by searching the maximum value of the Fourier spectrum.Then,the original vibration signal is decomposed into a series of mode functions by the alternating multiplier method,whose center frequencies are initialized at the estimated resonance frequencies.After that,the Hilbert transform is performed on the decomposition results to obtain the envelope spectrum of each mode function,and the fault characteristic frequency ratio(FCFR)is calculated.Finally,the mode function with the largest FCFR is selected as the effective component for envelope analysis to realize gearbox fault identification.Extensive experiments conducted on a 3 MW wind turbine transmission system validate the effectiveness and superiority of the proposed FIVMD method.

关 键 词：风电机组 状态监测 齿轮箱 故障诊断 变分模态分解 

分 类 号：TK83[动力工程及工程热物理—流体机械及工程]