Research On Rotor Fault Diagnosis Of Squirrel Cage Induction Motor

In the process of industrial development, the production is becoming automated, integrated, intelligent direction increasingly, as the main power transmission and transformation equipment the squirrel-cage motor is widely used in various types of production or control system, with the advantages of it’s simple structure and convenient control, however, an air-gap eccentricity or broken rotor bars faults may be occurred under the condition that the motor frequently started and braked or uninterrupted operation. The occurrence of these faults may cause the whole system ran unsteadily or trigger greater system faults, which not only affect the normal production and in severe cases can cause huge economic losses or endanger personal safety of the workers, therefore, research on motor fault diagnosis has become a research focus and has important meaning in avoiding catastrophic accidents and huge economic losses. With the method of signal processing and digital signal processing devices constantly upgrading, the motor fault diagnosis technology is upgrading too. The mainly faults in squirrel cage induction motor contain broken rotor bar, broken ring, air-gap eccentricity, motor bearing fault and stator winding inter-turn short circuit, we design an on-line monitoring system which based on RELAX algorithm and DSP in squirrel cage induction motor fault diagnosis by means of analyzing the corresponding characteristic frequency components which separate from the stator current signal to achieve the purpose that identification and diagnosis it’s faults. We use TI TMS320C6713high-speed DSP as the data acquisition and processing core of the whole system, the stator current signal is collected by the SP1T200C5V6current sensor, and then through the A/D conversion chip MAX1270the stator current signal which collected by the current sensor will converts into digital signals and transmits to the DSP processing. The hardware circuit design of the whole system is simple, anti-interference ability, high stability and reliability.The RELAX algorithm is based on frequency domain signal relaxation algorithm, is widely used in separating and extracting the characteristic signal from the source signals, compared with other algorithms it has very high recognition performance. First we use MATLAB simulation platform analysis weather the RELAX algorithm is feasible and the results shows it is really effective, then we write programs for the algorithm in the DSP development environment CCS3.3and download the programs into the DSP successfully, realize the purpose that separating the characteristic fault signal from the motor’s stator current signals. In the end the fault feature signals is sent to the host PC by RS232serial port, the monitoring software in the host PC is developed based on LabVIEW software development platform,by call the organizational competitive development of the neural network program which design by MATLAB development environment to complete the final classification and diagnosis of fault signal.