Industrial Servo Motor Yaskawa Electric SERVO MOTOR 200V 3000/min SGM-02A312 New in box

Thus, high costs associated with equipment to emulate the faults or destructive tests to generate datasets to train this method are not involved. The second advantage is related to scalability of the monitoring process. The signatures for the training and monitoring stages are normalized in amplitude. However, the signatures of the monitoring stage are not only normalized in amplitude, but also in frequency. This normalization in frequency of the signatures of the monitoring stage is a function of the signatures of the training stage. Thus, the signatures from the training and monitoring stages for the same motor operating condition have similar amplitude and frequency. These signatures with similar amplitude and frequency for the same motor operating condition are essential in the monitoring stage to yield high level of motor fault monitoring accuracy. Accordingly, the training and monitoring stages yield signatures that are independent of motor rated power, number of poles, level of load torque, and operating frequency of the real motor that is being monitored.

 

Thus, this method constitutes a powerful tool for induction motor fault monitoring. This is demonstrated and verified by the experimental results given in Chapter 5 of this thesis.