Research On Current Features For Motor Bearing Fault Based On DSP

Induction motors play an important role in the field of Agriculture and industrial productions because of their low costs, simple mechanism and high reliability. Once motors break down, giving rise to unplanned outage during the production, it would be at enormous economic cost. Bearing fault is common in motors failure, so it is of great significance on on-line fault diagnosis of machine to study the mechanism and features of Induction motor bearing fault.This paper works out the relationship between stator current wave component and load torque ripple component in bearing fault based on phase modulation theory put forward by Blodt. When the frequency of bearing fault is lower, the related stator current component is higher and the amplitude features of bearing fault are more obvious in spectrum analysis.In this paper, the identifiability of bearing fault to current diagnosis is increased by reducing the frequency of bearing fault. Induction motor model is built on Matlab software to do simulated test. The simulation results verify the relationship between the frequency of bearing fault and stator current amplitude features, meanwhile they verify stator current feature frequency relation deduced from theory. DSP is highly integrated, digital, high-speed computing and of rich resources on both internal and external sets. Consequently, this paper writes the program of SPWM and corresponding interrupt program on the basis of DSP platform hardware circuit of bearing fault current detection. Carrier wave ratio N is changed by the program to reduce the frequency, at the same time lowers voltage sharply, which makes ratio of voltage to frequency U/f less than normal. In this way, the output frequency of DSP testing platform can be changed. Recording the stator current signals on different Fault vibration frequencies, this paper analyzes current signal spectrum. Analysis results show that when the power supply frequency or frequency of fault vibration is reduced, stator current feature frequency is easier to be recognized.Based on results of the software simulation and physical experiment of drilling bearing outer race to verify the theories, this paper draws a conclusion that the methods mentioned before is practicable and effective in bearing fault detection.