模糊减法聚类在机组设备故障诊断的应用

Application of Fuzzy Subtractive Clustering in Fault Diagnosis of Unit Equipment

大型复杂机械系统的机组设备的故障诊断时通过提取故障信号特征,对特征数据采用聚类处理,实现故障识别。数据聚类是故障诊断的基础,传统方法中,采用模糊C均值聚类进行复杂机组设备故障分类识别和诊断,模糊C均值聚类对初始值和噪声极为敏感,导致故障识别率不高。提出一种基于模糊减法聚类算法的复杂机组设备故障诊断方法。首先分析了复杂机组设备故障诊断原理,提取复杂机组设备的振动信息作为原始数据,以此数据为基础进行信号模型构建,结合现代信号处理技术,采用功率谱估计实现对振动信号的故障特征提取,以功率谱为故障特征,采用模糊减法聚类算法进行数据分类识别,实现故障诊断。仿真实验进行了性能验证,仿真结果表明,采用该算法能有效提高复杂机组设备的故障识别率,具有较好的故障诊断性能。

The fault diagnosis of the equipment of the large complex mechanical system is characterized by extracting the fault signal. The feature data is processed by clustering, and the fault identification is realized. Data clustering is the foundation of fault diagnosis, in the traditional method, the fuzzy c-means clustering is used for the fault classification and diagnosis of complex equipment, fuzzy c-means clustering is very sensitive to the initial value and the noise, which causes that the fault recognition rate is not very high. A fault diagnosis method based on fuzzy subtractive clustering algorithm is proposed. First, the fault diagnosis principle of complex equipment is analyzed, and the vibration information of complex equipment is extracted as the original data, based on this data, the signal model is constructed. Combined with modem signal processing technologies, the power spectrum estimation is used so as to realize the fault feature extraction of vibration signals and its power spectrum is used as fault feature, the fuzzy subtractive clustering algorithm is used to classify the data to achieve fault diagnosis. Simulation experiments are carried out and the results show that the proposed algorithm can effectively improve the fault identification rate of complex units, and has good performance of fault diagnosis.