摘要
为解决风力发电机组故障轴承振动信号有效信息提取困难问题,提出一种基于变分模态分解(VMD)的多特征量风电机组轴承故障诊断法,首先利用VMD将振动信号分解为若干模态函数,对该组分量采用从时域指标、AR模型参数矩阵奇异值和能量熵三个角度的变化中提取和构造多特征量矩阵,输入支持向量机建立故障程度多分类预测模型,优化核参数和惩罚参数取得轴承故障最佳预测精度。实验验证所提方法是一种可行的风电机组轴承故障诊断法,能够有效提取轴承故障信息,提高轴承故障诊断率。
In order to solve the problem of difficult information extraction of fault vibration signals of wind turbine generators, a multi-feature wind turbine bearing fault diagnosis method based on Variational Mode Decomposition （VMD） was proposed. Firstly, the vibration signal was decomposed into several modal functions by using VMD, the multi-feature matrix was extracted and constructed from the changes of the time domain index, the AR model parameter matrix singular value and the energy entropy. The input support vector machine was used to establish the fault degree multi-classification prediction model. The nuclear parameters and penalty parameters were optimized to obtain the best prediction accuracy of bearing faults. The experiment verifies that this method is a feasible wind turbine bearing fault diagnosis method, which can effectively extract bearing fault information and improve bearing fault diagnosis rate.
作者
张瑶
张宏立
ZHANG Yao;ZHANG Hong-li(College of Electric Engineering,Xinjiang University,Urumqi Xinjiang 830047,China)
出处
《计算机仿真》
北大核心
2018年第9期98-102,共5页
Computer Simulation
基金
国家自然科学基金(51575469)
