Trend analysis combined with expert system applied to real-time monitoring of motor running status and warning of abnormal status

Chang-Ming Lin, Sheng-Fuu Lin

Energy saving, Fault diagnosis, Motor vibration, Percent tolerance pre-control chart, Process.

With the advancement of technology, industrial automation makes power equipment play an indispensable role in the industry. Power equipment is an indispensable source of power for manufacturing. Therefore, if the equipment is under high pressure, high temperature and high load for a long time, if there is an unwarned shutdown, it will cause huge economic losses. Although regular maintenance is used for maintenance work, it can reduce the occurrence of unwarned downtime, but it also consumes a lot of waste of human and material resources. Therefore, if you can know the real-time running status of the power equipment and the possible causes of the fault, early maintenance personnel can perform machine maintenance and troubleshooting to help improve the reliability of the system. Predictive maintenance is a kind of pre-warning and planning correction behavior. Through the analysis of the data, it can detect the initial abnormal reduction of the equipments additional energy consumption caused by the failure and improve the maintenance efficiency. According to the U.S. Department of Energy, Operations and Maintenance Best Practices, it can save about 5%~20% energy. In this study, the acceleration amplitude is used as the basis for instant detection. Collect the acceleration amplitude of the motor in the normal running state, and count the acceleration amplitude threshold of the motor in the normal running state. Monitor the amplitude of acceleration in real time to observe the running trend of the motor. When the running trend exceeds the normal range, the vibration spectrum characteristics combined with the experts rule are used to determine the possible abnormal types of the motor.

https://dx.doi.org/10.31873/IJEAS.7.05.22