A Fault Recognition Dataset for Eccentricity Faults in a Rotor-Eccentric Induction Motor

A fault recognition dataset generated from a Magnetic Equivalent Circuit Model [1, 2] of a Cantoni 2SIE80-2B with a wye configuration at 230V and 50kHz and applied torque varying between 0Nm and 1.3Nm. The dataset contains 31,920 instances, divided into 25,530 for training and 6,390 for testing. It includes the stator currents, applied phase voltages, rotor speed, and rotor angle, all sampled at 5000Hz with windows of 256 samples. Check the README.txt file for further details about this dataset.
<br><br>
[1] P. Desenfans, Z. Gong, D. Vanoost, D. Pissoort, A Python Magnetic Equivalent Circuit Model for Rapid Data Generation of Induction Motors in Faulty Conditions. KU Leuven, Bruges, 2024.
<br>
[2] P. Desenfans, Z. Gong, D. Vanoost, K. Gryllias, J. Boydens, D. Pissoort, The influence of the unbalanced magnetic pull on fault-induced rotor eccentricity in induction motors. Journal of Vibration and Control. 2023.