Application of principal component analysis for the fault detection and diagnosis of active magnetic bearings Online publication date: Thu, 01-Feb-2018
by Anand S. Reddy; Praveen Kumar Agarwal; Satish Chand
International Journal of Advanced Mechatronic Systems (IJAMECHS), Vol. 7, No. 4, 2017
Abstract: Active magnetic bearing (AMB) is an advanced mechatronic system. For achieving desired operating characteristics of AMB system, all of its components (position sensors, actuators or electromagnets, etc.) should work satisfactorily without any faults. Fault or failure of any component may result into destructive rotor dynamics and sometimes catastrophic failure of the entire system. Therefore, it is essential to know the occurrence of fault and to identify faulty component immediately for avoiding the failure of entire system. This can be achieved by online fault detection and diagnosis (FDD) of AMB system. Various model based and redundant component based methodologies have been proposed for FDD of sensors and actuators of AMBs. Model based FDD methods require complex mathematical modelling and have higher chances of subjected to modelling errors. Redundant sensors and actuators based methods incur additional cost and also require additional space for installation. Therefore, in the present work, simulation data driven principal component analysis (PCA) based FDD with statistical analysis is designed for detecting and diagnosing faults in position sensors and actuators. Various types of faults such as bias, multiplicative and noise addition are successfully diagnosed.
Online publication date: Thu, 01-Feb-2018
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Advanced Mechatronic Systems (IJAMECHS):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email email@example.com