Marine fire pump motor bearings fault feature enhancement and diagnosis based on adaptive SSA and improved TEO

Objectives The working environment of marine fire pump motor bearings is complex with low fault diagnosis accuracy. To address these issues, this study proposes a fault feature enhancement and diagnosis method for marine fire pump motor bearings based on adaptive steady-state subspace analysis (SSA) and improved Teager energy operator (TEO).

Methods  First, the traditional SSA algorithm is optimized to adaptively determine the dimensionality of the Hankel matrix by the false nearest neighbor method, and non-stationary signals with the best fault features are extracted from the vibration signal through kurtosis. Second, by improving the TEO algorithm, the proportion of fault feature information in the vibration signals is increased, fault features are enhanced and faults are diagnosed. Finally, the effectiveness of the method is verified through simulation and engineering experiments.

Results The proposed method can accurately distinguish the fault characteristic frequency and harmonics of bearings, and accurately diagnose bearing faults.

Conclusions The results of this study can provide references for the fault diagnosis of marine pump motor bearings.