Application of Impulse Demodulation Analysis in the Abnormal Noise Fault Diagnosis of Robot Joint

In view of the obvious abnormal noise in the joint of an industrial robot at low speed, the joint vibration signal is collected and analyzed by impulse demodulation method. Firstly, in order to improve signal-to-noise ratio effectively, the fault signal is detected and demodulated by vibration waveform envelope method. Then, the filtered signal is processed by FFT algorithm to extract the characteristic frequency caused by low frequency impulse. Finally, it is found that there are abnormal periodic shocks in the vibration signal at low speed. The shock modulation frequency is the motor rotation frequency. At high speed, abnormal noise and abnormal shocks disappear. It is inferred that the abnormal sound is caused by the poor meshing of the joint gear caused by the abnormal external load. The monitoring results show that the gear and bearing of the joint reducer are in normal condition, so there is no need to check the robot disassembly, and a large overhaul work is avoided.