Detection of rotor faults in induction motor in non-stationary conditions using the matrix pencil method

Increasing the economic efficiency of production by reducing unplanned downtime and optimizing the maintenance of industrial equipment requires the development of diagnostic methods for such key equipment components as induction motors. In turn, the operating mode of the electric motor has a significant impact on the accuracy and reliability of diagnostics. In particular, the widely used method of motor current signature analysis (MCSA) has a number of limitations when the motor is operating at variable speed or load. Therefore, it is necessary to investigate frequency analysis methods that can extract motor defect signs from non-stationary current signals. Aim of the study: to apply matrix pencil method to motor current signals to extract signs of defect in the rotor bar of an induction motor under non-stationary mode of operation.