Rotor Eccentricity Quantitative Characterization Based on Physics-Informed Adversarial Network and Health Condition Data Only

Rotor eccentricity can lead motor damage, so it is necessary to obtain the quantitative characterization of rotor eccentricity for motor health management or suppression. Neural network is an effective method for eccentricity quantitative characterization, and a large amount of data is needed to tra...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on industrial electronics (1982) 2024-07, Vol.71 (7), p.1-15
Main Authors: Wang, Haowen, Sun, Wei, Jiang, Dong, Liu, Zicheng, Qu, Ronghai
Format: Article
Language:English
Subjects:
Data models
Eccentricity
Induction motors
Industrial applications
Mathematical models
Neural networks
Permanent magnet motors
physics-informed adversarial network (PIAN)
quantitative characterization
Rotors
Synchronous motors
Training
Windings
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Rotor eccentricity can lead motor damage, so it is necessary to obtain the quantitative characterization of rotor eccentricity for motor health management or suppression. Neural network is an effective method for eccentricity quantitative characterization, and a large amount of data is needed to train the network. However, different eccentricity data are hard to collect in industrial application. To solve this problem, the physics-informed adversarial network is proposed in this article for eccentricity quantitative characterization with only health data. The network learns physical law from a physics-informed mathematical model of a motor, and then the model data and actual health data are combined for adversarial training. The trained network can quantitatively characterize the eccentricity. Simulation and experiment results show that the proposed method has a good performance even if the errors of motor parameters exist.
ISSN:0278-0046
1557-9948
DOI:10.1109/TIE.2023.3306397