Broken bar detection based on measurement of induction motor one phase start-up current
2011
Matić Dragan | Bugarski Vladimir | Kamenko Ilija | Nikolić Perica
The paper presents the method of the broken bar detection based on analysis of an induction motor start-up current for no load case. Motor Current Signature Analysis (MCSA) is far the most popular approach to detect induction motor failures these days due to its benefits. The transient analysis is conducted by calculating power spectrum density (PSD) via Sort Time Fourier Transformation (STFT). Discriminative feature is obtained by calculating skenwness statistical parameter for given PSD for total amount of power per segment. It is shown that values for chosen features differ for healthy and broken bar rotor condition for a wide load range. Based on the proposed features estimation of rotor condition can be made. Study is done on real motor data collected from laboratory and industrial environment.
显示更多 [+] 显示较少 [-]