Monitorowanie napędów elektrycznych z silnikami indukcyjnymi za pomocą wektora przestrzennego prądów stojana

In the important industrial processes using electrical drives of high power, the on-line monitoring and diagnosis systems are implemented, because of economical reasons. This paper deals with one of such monitoring methods, which uses the space vector of stator current for diagnosis purposes. The proposed method connects the simplicity of well-known method of motor current signature analysis (MCSA) with simultaneous increase of the accuracy of spectral analysis. The application of space current vector method (Park's vector method) for the detection of broken rotor bars, drive motor misalignment and stator windings asymmetry of the induction motor was demonstrated. The laboratory benchmark for the induction motor faults testing was described. This laboratory rig was equipped with the automatic fault detector realized in digital signal processor based on TMS320C31. Basic failures of the induction motor were tested using space vector method. Frequency spectra of Park's vector modulus of the faulted rotor, stator with short-circuit turns and drive misalignment were presented. The proposed method was evaluated from the point of view of the development of neural network based fault detectors.