Research On The Current Features For Motor Bearing Faults Based On FPGA

The Induction motor has simple construction and wide application fields and its sustained, stable and reliable operation are important for the safety of personal and system. Once the motor faults occur, the motor can be damaged and the system may be paralyzed. Bearing fault is the main fault type for motor faults, so the study on the bearing fault mechanism as well as the current character and the evaluation of motor operating state can be helpful for bearing fault detection. Bearing fault detection is the key to make sure that the motor and system can be running effectively.Based on the torque fluctuation model proposed by Blodt, the paper proposes that the IF (intermediate frequency) sinusoidal signal can be injected into the PWM and the IF sinusoidal signal modulates with the bearing fault vibration frequency. The result shows that some new bearing fault characteristic frequencies are produced in the IF, then the bearing fault characteristic frequency can avoid the high noise background in the low frequency, which can improve the characteristic frequency recognition. The paper analyzes the influence on the inverter voltage when the IF sinusoidal signal are injected theoretically, and gives stator current characteristic frequency expression when the bearing fault occurs after the IF sinusoidal signal are injected into the PWM. In addition, the paper analyzed the factors affecting the current characteristic frequency identification after injected IF sinusoidal signal.The paper designs the software and hardware for the motor drive system based on FPGA with the IF sinusoidal signal is injected into the PWM signal. The system makes the IF sinusoidal signal injection into the PWM signal come true, which can provide the related experiment environment.Due to the torque fluctuation caused by bearing fault is small, the amplitude of the characteristic frequency reflected in the stator current is smaller, so it is difficult to find the current characteristic frequency feature. In this case, bearing fault simulated platform is used to build the bearing fault detection platform to do research on the current characteristic frequency when the IF sinusoidal signal injected into the PWM signal. The results show that current characteristic frequency expression is valid, at the same time it is proved that the method of IF sinusoidal signal injection is effective to bearing fault detection and recognition.