摘要
针对Weibull分布模型不能有效的提取具有非线性、非平稳性的滚动轴承振动信号信息,提出了基于广义Weibull分布模型参数的滚动轴承故障特征提取方法。首先,将采集的4种运行状态的滚动轴承振动信号进行Hilbert变换,求得能表征4种运行状态下振动信号幅值和频率变化的Hilbert包络信息;然后对Hilbert包络信息建立两参数Weibull分布模型,利用最小二乘法估计模型的尺度参数和形状参数;最后将估计出的尺度参数和形状参数作为表征滚动轴承运行状态的特征信息输入SVM分类器进行模式识别和故障诊断。通过Matlab仿真实验表明,提出的特征提取方法能快速、准确地提取滚动轴承的特征信息,对工程实践具有一定的指导意义。
Since the Weibull distribution model,we can not extract the vibration signal information of rolling bearing with nonlinearity and nonstationarity. In this paper,we propose a method for extracting fault feature of rolling bearing based on generalized Weibull distribution model. Firstly,the Hilbert transform of four kinds of running bearing vibration signals is obtained by Hilbert transform,and the Hilbert envelope information can be obtained to represent the amplitude and frequency of the vibration signal in the four operating states. Then,the two-parameter Weibull distribution is established for the Hilbert envelope information. The model is used to estimate the scale parameters and shape parameters of the model by using the least squares method. Finally,the estimated scale parameters and shape parameters are input to the SVM classifier as the characteristic information of the running state of the rolling bearing for pattern recognition and fault diagnosis. The simulation results show that the feature extraction method proposed in this paper can quickly and accurately extract the characteristic information of rolling bearing,which has certain guiding significance to engineering practice.
作者
姜海燕
林波
JIANG Haiyan;LIN Bo(College of Railway Power and Electrical, Hunan Railway Professional Technology College, Zhuzhou 412001, CHN;School of Electrical and Automation Engineering, East China Jiaotong University, Nanchang 330013, CHN)
出处
《制造技术与机床》
北大核心
2018年第6期81-86,共6页
Manufacturing Technology & Machine Tool
基金
2016年湖南省教育厅科学研究项目(16B176)
