A time-domain fault detection method based on an electrical machine stator current measurement for planetary gear-sets

Fault diagnosis of geared drive-train systems is usually based on vibration monitoring. However, such vibration based techniques are difficult to implement in planetary gearboxes due to the complex nature of measured vibration spectrum. Motor current signal analysis (MCSA) provides an alternative and non-intrusive way to detect mechanical faults through electrical signatures. In this paper, a new time-domain fault detection algorithm is presented for the detection of planetary gear faults using electrical machine stator current signals. This time-domain fault detection method combines fast dynamic time warping (DTW) and correlated kurtosis techniques to process the current signals data to detect and identify damaged planetary gear and its position. Fast DTW is employed to highlight the sideband patterns resulting from tooth damage by the introduction of an estimated reference signal that has the same frequency as the gear mesh frequency. Correlated kurtosis (CK) takes advantages of the periodicity of the geared faults; it is used to identify the position of the damaged gear tooth in the planetary gear-set. This method is later applied to simulated current signals generated from a lumped parameter model of planetary gearbox driving a permanent magnet synchronous generator to evaluate its performance. The simulated results demonstrate the effectiveness of the proposed time-domain approach to detect faults in planetary gear-sets based on the electrical stator current signal.