二维曲线相似与统计学原理在旋转设备故障识别中的应用

在机械设备状态检修中使用振动故障诊断已经相当普及。目前普遍采用设置阈值或数值的变化趋势实现故障预警,再通过大量的故障诊断算法实现故障识别。提出一种新的基于几何相似和统计学原理的故障频谱相似度算法,可以简单有效快速的实现故障识别。

Application of Two-dimentional Curve Similarity and Statistics Principles in Identifysing Rotating Equipment Faults

The vibration fault diagnosis of mechanical equipment maintenance in the state has been popularly used. The trend now is widely used to set the threshold value to achieve fault or warning. The fault identification through a lot of fault diagnosis algorithm is realized. This paper presents a new statistical theory based on geometric similarity and the similarity of the fault spectrum algorithm can simply and effectively achieve rapid fault identification.