基于判别稀疏编码的轴承故障诊断方法  

Bearing Fault Diagnosis Method Based on Discriminative Sparse Coding

作　　者：王鹏飞[1] 王新晴[1] 曹蕾[2] 王云龙[1] 李艳峰[1] 

机构地区：[1]解放军理工大学,江苏南京210007 [2]苏州市公安局相城分局,江苏苏州215131

出　　处：《仪表技术与传感器》2016年第8期77-80,共4页Instrument Technique and Sensor

摘　　要：为解决轴承故障诊断中故障信号特征难以提取、不同故障程度间信号特征相近难以区分的问题,提出了基于判别稀疏编码的轴承故障诊断方法:在稀疏编码框架下,引入Fisher判别准则,增强不同类别故障字典的判别性,并基于重构误差,在频域上对故障信号进行处理。实验表明:与其他方法相比,该方案有效提高了轴承故障诊断的准确率,并具有较好的稳定性。According to the fact that the features of bearing fault signal were difficult to extract,and they were difficult to be distinguished between different degrees of failure,the method of bearing fault diagnosis based on discriminative sparse coding was proposed.It’s a sparse coding framework involving a Fisher discriminative criterion to increase identification ability for different categories of fault dictionaries.The signal types were identified in the frequency domain based on the reconstruction error.Verified by experiment and compared with other methods,the method based on discriminative sparse coding can improve the accuracy and stability of the bearing fault diagnosis.

关 键 词：稀疏编码 判别性字典 轴承故障诊断 FISHER判别准则 

分 类 号：TH113[机械工程—机械设计及理论]