基于时频图像特征提取的状态识别方法研究与应用

PDF(1495 KB)

振动与冲击 ›› 2010, Vol. 29 ›› Issue (7) : 184-188.

论文

基于时频图像特征提取的状态识别方法研究与应用

李宏坤1,2;周帅1;黄文宗1

作者信息 +

Investigation on Time-Frequency Image Feature Extraction for Machine Condition Classification and Application

Li Hong-Kun1,2;Zhou Shuai1;Huang Wenzong1

Author information +

文章历史 +

摘要

本文介绍一种基于时频图像进行设备状态识别的新方法。论述了对振动信号应用Hilbert谱构建二维时频图像，研究时频图像三维重心与信息熵的特征提取方法构造特征向量，采用支持向量机进行设备的状态识别。并以滚动轴承不同状态的识别为例，验证方法的有效性。为提高滚动轴承振动信号特征提取的可靠性，本文采用多循环平均方法减小循环波动性影响。研究表明此方法能够提高设备故障诊断与状态识别的准确性。

Abstract

In this paper, a new machine condition pattern recognition method is introduced. Hilbert spectrum is used to construct time-frequency image for vibration signal analysis. Time frequency image gravity centre and information entropy are investigated to construct feature vector. Support vector machine is used as classification tool for pattern recognition. Rolling bearing pattern recognition is as an example to verify effectiveness of this method in the research. Multi-cycle synchronous average is used to reduce cyclostationary effect between different cycles to improve the recognition accuracy. According to the result analysis, it can be concluded that this method will contribute the development of machine fault diagnosis and pattern recognition.

导出引用

李宏坤;周帅;黄文宗 . 基于时频图像特征提取的状态识别方法研究与应用[J]. 振动与冲击, 2010, 29(7): 184-188

Li Hong-Kun;Zhou Shuai;Huang Wenzong. Investigation on Time-Frequency Image Feature Extraction for Machine Condition Classification and Application[J]. Journal of Vibration and Shock, 2010, 29(7): 184-188