Fuzzy-Bayes Fault Isolator Design for BLDC Motor Fault Diagnosis

  • Published : 2004.09.01

Abstract

To improve fault isolation performance of the Bayes isolator, this paper proposes the Fuzzy-Bayes isolator, which uses the Fuzzy-Bayes classifier as a fault isolator. The Fuzzy-Bayes classifier is composed of the Bayes classifier and weighting factor, which is determined by fuzzy inference logic. The Mahalanobis distance derivative is mapped to the weighting factor by fuzzy inference logic. The Fuzzy-Bayes fault isolator is designed for the BLDC motor fault diagnosis system. Fault isolation performance is evaluated by the experiments. The research results indicate that the Fuzzy-Bayes fault isolator improves fault isolation performance and that it can reduce the transition region chattering that is occurred when the fault is injected. In the experiment, chattering is reduced by about half that of the Bayes classifier's.

Keywords

References

  1. Ph.D. thesis, Dankook Univ. Fault Detection and Isolation Schemes Using State Observers and Parity Equations T. G. Park
  2. IEEE Trans. on Industrial Electronics v.47 no.5 Fault detection and diagnosis of permanent-magnet DC motor based on parameter estimation and neural network X.-Q. Liu;H.-Y. Zhang;J. Liu;J. Yang https://doi.org/10.1109/41.873210
  3. IEEE Trans. on Industrial Electronics v.47 no.5 Application of model-based fault detection to a brushless DC motor O. Moseler;R. Isermann https://doi.org/10.1109/41.873209
  4. Fault Diagnosis Using Parameter Estimation and Supervised Classification Techniques O. Norretranders
  5. International Journal of Control, Automation, and Systems v.1 no.3 Serial communication-based fault diagnosis of a BLDC motor using Bayes classifier S. H. Suh;K. J. Woo
  6. Pattern Classification R. O. Duda;P. E. Hart;D. G. Stork
  7. Fuzzy Control K. M. Passino;S. Yurkovich
  8. Model-based fault detection for a brushless DC motor using parameter estimation O. Moseler;R. Isermann
  9. Proc. of the American Control Conference Model-based fault diagnosis of an actuator system driven by the brushless DC motor O. Moseler;D. Juricic;A. Rakar;N. Muller
  10. System Identification Theory for the User L. Ljung
  11. Journal of Control, Automation and Systems Engineering v.5 no.4 An improved new RLS algorithm with forgetting factor of Erlang function for system identification J.-W. Seok;K.-S. Choi;J.-S. Lee;S.-W. Cho
  12. Electronic Machinery and Transformers I. L. Kosow