Download
Abstract

Active magnetic bearings are complex mechatronic systems that consist of mechanical, electrical, and software parts, unlike classical rolling bearings. Given the complexity of this type of system, fault detection is a critical process. This paper presents a new and easy way to detect faults based on the use of a fault dictionary and Machine Learning. In particular, the dictionary was built starting from fault signatures consisting of images obtained from the signals available in the sys-tem. Subsequently, a convolutional neural network was trained to recognize such fault signature images. This work concentrates on recognizing the most frequent electrical faults that typically af-fect position sensors and actuators. This new method permits, in a computationally convenient way, that can be implemented in real time, to determine which component has failed and what kind of failure has occurred. Therefore, this fault identification system allows for determining, in the event of a fault, which countermeasure to adopt in order to enhance the reliability of the sys-tem. The performance of the proposed method is assessed by means of a case study concerning a real turbomachine supported by two active magnetic bearings for the oil and gas field. Seventeen fault classes have been considered and the neural network fault classifier reached an accuracy of 93% on the test dataset.

Citation

Donati, G., Basso, M., Manduzio, G. A., Mugnaini, M., Pecorella, T., & Camerota, C. (2023). A Convolutional Neural Network for Electrical Fault Recognition in Active Magnetic Bearing Systems. Sensors, 23(16), 7023. https://doi.org/10.3390/s23167023


@Article{s23167023,
AUTHOR = {Donati, Giovanni and Basso, Michele and Manduzio, Graziano A. and Mugnaini, Marco and Pecorella, Tommaso and Camerota, Chiara},
TITLE = {A Convolutional Neural Network for Electrical Fault Recognition in Active Magnetic Bearing Systems},
JOURNAL = {Sensors},
VOLUME = {23},
YEAR = {2023},
NUMBER = {16},
ARTICLE-NUMBER = {7023},
URL = {https://www.mdpi.com/1424-8220/23/16/7023},
PubMedID = {37631560},
ISSN = {1424-8220},
DOI = {10.3390/s23167023}
}