In-line identification of oil debris signals: an adaptive subband filtering approach

The oil debris monitor (ODM) is a device used to examine lubricant oil conditions. However, the ODM is prone to both background noise and vibration interferences, thereby causing false alarms and also limiting its ability in detecting fine particles. This paper focuses on the enhancement of the ODM performance. This is achieved by a two-stage de-noising scheme. In the first stage, a wavelet-based adaptive subband filtering technique is applied to remove the vibration-related interferences. The outputs of the adaptive filters are then thresholded in the second stage to remove the background noise mainly caused by the wiring and measurement system flaws. The proposed approach has been validated using both simulated and experimental data.