Servo DC Motor Fault Diagnosis Driven by Multi-source Heterogeneous Data

According to the requirement of the on-line fault diagnosis of the servo DC motors used in TT&C antenna, which is applicated in aerospace exploration, noise reduction and power spectrum analysis with wavelet packet are studied in this paper. The ability of feature extraction and pattern recognition are the characteristics of wavelet packet noise reduction, which are applied to analyze the motor's current signals in different branches, thus the characteristic parameters of the fault motors can be shown clearly. We proposed a servo DC motor fault diagnosis model based on informatization theory, which improves the reliability of fault diagnosis result, and the accuracy of data screening increases from 70.5% to 92%. The result shows that the fault diagnosis model can trigger alarm accurately, it is also effective and superior in fault location.