Demagnetization Quantification of PMSM Based on Support Vector Regression

Demagnetization is one typical fault for Permanent Magnet Synchronous Motor (PMSM). The control performance, efficiency and reliability of PMSM depend greatly on the healthy condition of the rotor permanent magnet, the quantification of demagnetization therefore can be regarded as an performance ind...

Full description

Saved in:
Bibliographic Details
Published in2018 Prognostics and System Health Management Conference (PHM-Chongqing) pp. 619 - 623
Main Authors Dou, Renjie, Song, Fangzhen, Liu, Haining, Men, Xiuhua
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2018
Subjects
Analytical models
Circuit faults
Demagnetization
demagnetization fault
Mathematical model
Matlab/Simulink
Permanent magnet motors
permanent magnet synchronous motor (PMSM)
Support vector machines
support vector regression
Torque
Online AccessGet full text

Cover

More Information
Summary:Demagnetization is one typical fault for Permanent Magnet Synchronous Motor (PMSM). The control performance, efficiency and reliability of PMSM depend greatly on the healthy condition of the rotor permanent magnet, the quantification of demagnetization therefore can be regarded as an performance index for PMSM. Data-driven fault diagnosis does not rely on the prior knowledge of the system, and uses the data mining technology to obtain the feature information to express the normal and the fault mode of the system. In this paper, a Matlab/Simulink model is established to simulate the different levels of demagnetization for a PMSM motor, current and torque signals are obtained, and then features are extracted. Finally, the Support Vector Regression (SVR) is employed to map the feature vectors into the demagnetization levels. With this model, time and fund investment can be saved to obtain the dynamic characteristics of PMSM, and diagnostic modeling for PMSM condition monitoring would also become straightforward.
ISSN:2166-5656
DOI:10.1109/PHM-Chongqing.2018.00111