Electrostatic monitoring of rolling bearing with multi-sensor fusion under variable operating conditions

Abstract

The friction and wear condition of rolling bearing can be detected and monitored by the novel electrostatic sensing technique. It has been verified under stable operating conditions, while the change of conditions impact a lot for the electrostatic original signals. This paper introduced a new method called moving window local outlier factor (MWLOF) to process electrostatic monitoring signals of rolling bearing under variable operating conditions. Compared with traditional features, the extracted features can reduce the impact and accurately reflect the wear condition of the bearings in the load and accelerated life tests. It can detect early faults earlier and has a better sensitivity and performance degradation trend than conventional techniques, which leads to a better industrial application in future.