Research On Detection Method For Motor Bearing Defect Based On FPGA And DSP

Induction motors are widely used in residential and industrial production due to their good running performance,low price,simple structure and wide applicability.As an important part of the induction motor,the bearing has a high probability of damage,which can easily lead to operational failure of the induction motor.Therefore,early diagnosis of motor bearing defect is of great significance.At present,most of the research on defect diagnosis of bearing with stator current is based on the torque fluctuation model of bearing defect proposed by Blodt.However,in the actual current signal,it is often found that there are two or three times of frequency.The torque fluctuation model proposed by Blodt cannot explain this phenomenon very well.Therefore,in order to more accurately study the stator current characteristics of bearing defect,this paper analyzes the torque ripple caused by bearing defect based on dynamic theory.The simulation results show that the torque double pulse model caused by defect is more in line with the actual situation.Based on the torque double-pulse model,this paper re-derives the characteristic frequency representation of bearing defect in the stator current.It is found that there is not only one characteristic frequency in the stator current spectrum,but also high-order multiple frequency,with bearing defect.The change in the situation,the characteristic frequency multiplication in the current spectrum will also change accordingly.In order to visually observe the characteristic frequency of bearing defect,the current demodulation method is used to demodulate the current,so that the frequency doubling of the defect characteristic frequency can be observed directly in the current spectrum.In order to improve the signal-to-noise ratio of the defect characteristics,this paper adopts the operation mode of voltage regulation and speed regulation to reduce the proportion of sliding friction torque.This method can also make the defect characteristic frequency away from the fundamental frequency,thus improving the recognizability of the defect characteristic frequency.In the laboratory,an FPGA-based motor PWM experimental platform is built to control the frequency and voltage of the motor.The stator current measurement module is designed by using AD7606,and the stator current signal processing platform based on DSP is built.The sampled stator current signal is coherently demodulated and then analyzed.The experimental results show that the method based on this paper can effectively extract the multiples of the current spectrum,thus realizing the online monitoring of motor bearing defect.