摘要
针对传统故障诊断中,特征的有效提取依赖于降噪的效果,提出一种基于多域熵与模糊C均值聚类的故障诊断模型。采集设备运行过程中的振动信号,分别计算其小波包能量熵、功率谱熵和近似熵,其反映了振动信息在小波域、频域以及时域内的复杂程度。将其作为设备运行特征向量,通过模糊C均值聚类对设备状态进行识别。利用轴承故障实验和转子故障实验验证基于多域熵与FCM聚类的故障诊断模型,结果表明该方法地对故障类别以及故障程度的识别分类具有良好的效果。
The effective extraction of features depends on the effect of noise reduction in traditional fault diagnosis. A fault diagnosis model based on multi-domain entropy and fuzzy C-means clustering( FCM) is proposed. The vibration signals of working equipments are calculated respectively,and then their wavelet packet energy entropy,power spectrum entropy and approximate entropy are calculated respectively. Therefore the wavelet domain,frequency domain and time domain's complexity of vibration information are reflected and regarded as a feature vector of equipment operation. The device status is identified by fuzzy C-means clustering of the feature vector. Fault diagnosis model based on multi-domain entropy and FCMclustering is validated by the bearing fault and rotor fault experiments. The results showthat the method has good classification effect on the fault type and degree of fault.
出处
《组合机床与自动化加工技术》
北大核心
2016年第8期64-66,共3页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国家自然科学基金资助项目(51575055)
国家科技重大专项(2015ZX04001002)
关键词
多域熵
FCM聚类
故障诊断
multi-domain entropy
fuzzy C-means clustering(FCM)
fault diagnosis