Use of Neural Networks in Diagnostics of Rolling-Element Bearing of the Induction Motor

Bearing defect is statistically the most frequent cause of an induction motor fault. The research described in the paper utilized the phenomenon of the current change in the induction motor with bearing defect. Methods based on the analysis of the supplying current are particularly useful when it is impossible to install diagnostic devices directly on the motor. The presented method of rolling-element bearing diagnostics used indirect transformation, namely Clark transformation. It determines the vector of the spatial stator current based on instantaneous current measurements of the induction motor supply phases current. The analysis of the processed measurement data used multilayered, one-directional neural networks, which are particularly attractive due to their nonlinear structure and ability to learn. During the research 40 bearings: undamaged, with damages of three types and various degrees of fault extent, were used. The conducted research proves the efficiency of neural networks for detection and recognition of faults in induction motor bearings. In case of tests of the unknown state bearings, an efficiency approach to failure detection equaled 77%.