Fault Diagnosis of Control Moment Gyroscope Using Optimized Support Vector Machine

Control moment gyroscope is used as an attitude control system in satellites and its failure may results in mission failure. Fault diagnosis can prevent this if accompanied by an in-time remedial action. In this paper, a data-driven fault diagnosis model is developed using an optimized support vecto...

Full description

Saved in:
Bibliographic Details
Published in2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC) pp. 3111 - 3116
Main Authors Varvani Farahani, Hossein, Rahimi, Afshin
Format Conference Proceeding
LanguageEnglish
Published IEEE 11.10.2020
Subjects
control moment gyro
Fault diagnosis
Gyroscopes
Predictive models
Principal component analysis
Satellites
support vector machine
Support vector machine classification
Wavelet packets
Online AccessGet full text

Cover

More Information
Summary:Control moment gyroscope is used as an attitude control system in satellites and its failure may results in mission failure. Fault diagnosis can prevent this if accompanied by an in-time remedial action. In this paper, a data-driven fault diagnosis model is developed using an optimized support vector machine to diagnose multiple in-phase faults of the satellite control moment gyroscope. The wavelet packet transform is used for feature extraction. The principal component analysis is used to reduce the number of features. Grid search is used to find the optimum values for the model hyperparameters and cross-validation with five-folds is used to validate the final model. The results show that the proposed model can predict faults with 62% accuracy.
ISSN:2577-1655
DOI:10.1109/SMC42975.2020.9283402