基于统计学习的振动信号监测技术研究

Research on Sensitive Monitoring Based on Statistical Learning for Vibration Signals

介绍了一种集报警监测、统计学习以及历史数据存储为一体的灵敏监测技术。通过对历史数据的统计学习,该项技术能够自动确定各分频报警阈值,并在报警事件发生时触发存储机制,捕捉故障数据。此技术在在线状态监测系统S8000中发挥了相当重要的作用。

In order to monitor the rotating mechanical vibrations,after contriving to traditional monitoring methods,the paper introduced a new sensitive monitoring technology which was composed of multiple moni- toring targets,statistical learning ability and data storage.Through the statistical study of historical data,the technology can determine automatically the threshold value of various frequencies.As soon as the alarm event occurs,the system triggers the memory mechanism and captures breakdown data.This technology has played a very important role in the condition monitoring system S8000.