Zhiping Xie, Jianfeng Hong , Wei Wu, and Wenxiang Chen
Micro direct-current motor (MDCM), fault detection, multiple features, spectrum analysis, statistical pattern
This paper presents a new fault detection method for the micro direct-current motor (MDCM) that can be widely used to ensure the quality of outgoing products. The method is based on multiple features of the armature current in the frequency domain. The analytical expression for the armature current is obtained from the dynamic model of DC motor under no-load and steady-state operation conditions. The five multiple features extracted from the armature current of the MDCM are utilized as the fault detection indices. Using the method of conventional hypothesis testing, varieties of possible theoretical distributions of the multiple features are enumerated and compared. Then, statistical models of and threshold values for the multiple features are obtained. Using this method, faulty motors could be distinguished from healthy motors by comparing the multiple features of the tested motors with the threshold values. The proposed methodology was validated experimentally on the fault detector designed for an enterprise. The experimental results showed that the correct classification rate of faulty motors is 100% and the correct classification rate of healthy motors is 95.83%.
Important Links:
Go Back
