Fault detection of rotating machine elements by wideband spectral histogram analysis

Wideband spectral histogram 분석에 의한 회전기계 요소의 결함탐지

The use of vibration signal for fault detection inside a machine (or its part) is not new, and a number of signal processing techniques are available at the moment. A common problem for all such techniques is the high level of background signals which often made it very difficult to distinguish the desired signals. We developed a new algorithm to detect the minute faults of rotating part based on spectral histogram analysis. The technique has subsequently been applied to the bearings and has proved to be useful.