Research Of Asynchronous Motor Fault Detection System Based On DSP

Asynchronous motor has many advantages such as simple structure, low cost and so on, which lead to the widespread application in the modern industrial and agricultural production. As the main power source in the field of production, the normal work of the machine is very meaningful for ensuring the safe, efficient, agile, high quality and low energy consumption operation in the production. If there is something wrong with motor in the process of production, the production will be stopped and it will bring huge losses. So we should timely monitor the motor’s working condition, find the fault and repair it, which can prolong the service life of the motor, improve the production efficiency and guarantee personal safety. This article investigates the motor fault detection system based on DSP,which has very vital significance in the industrial and agricultural production.In view of the fault problems in the motor, the wavelet packet analysis method is used to test the fault signal, as the fault signal will contain a large number of sudden-changeable signal, we can obtain the fault information of the machine by signal detection and analysis.The Fourier transform is used to analyze the frequency domain of the signal in the traditional fault detection method, which only analyzes a certain frequency amplitude of the size and density. However, it cannot detect the singular point of time-domain signal,which only reflect the whole characteristics of the signal and doesn’t have the time-domain local characteristics. Wavelet packet analysis is put forward on the basis of the wavelet theory, which overcomes the shortcoming of the fourier transform and inherits the advantages of wavelet transform. According to the property of multi-resolution analysis of wavelet packet, the signal can be projected to the different frequency band to be dealt and compared. Because most of the fault signal is a nonlinear signal, this method can achieve the result of fault detection. In order to accurately extract the fault information in fault diagnosis, the signal noise reduction is necessary. This paper adopts the modified wavelet packet threshold method, which overcomes the shortcomings of traditional soft and hard threshold de-noising method, enhances the noise control precision, and improves the precision of fault diagnosis.This paper design the motor fault detection system platform with DSP processor, it collects the motor stator current signal through the hall current sensor. After the process of signal conditioning circuit and A/D, the signal data will be transmitted to DSP and dealt with by combining wavelet packet with BP neural network algorithm procedures. Upper computer receives signal via a serial port communication, and tests the fault condition of the asynchronous motor.Designing the monitoring interface based on LabVIEW platform, which we can observe the fault information of the motor. Through the test of the asynchronous motor fault diagnosis, it shows that the motor fault detection system is stable and reliable, good real-time and can accurately detect the common fault of asynchronous motor.