Broken Rotor Bar Detection Based on Steady-State Stray Flux Signals Using Triaxial Sensor with Random Positioning

• Published in: Sensors (Basel, Switzerland) Vol. 24; no. 10; p. 3080
• Main Authors: Zubčić, Marko, Pavić, Ivan, Matić, Petar, Polak, Adam
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 12-05-2024
• Subjects: Analysis; broken rotor bar detection; Fourier transforms; Induction electric motors; Investigations; Literature reviews; Magnetic fields; Methods; non-parametric test; parametric test; sensor random position; Sensors; Signal processing; squirrel cage induction motor; Wavelet transforms; non-parametric test; sensor random position; broken rotor bar detection; squirrel cage induction motor; parametric test
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s24103080

This paper investigates the detection of broken rotor bar in squirrel cage induction motors using a novel approach of randomly positioning a triaxial sensor over the motor surface. This study is conducted on two motors under laboratory conditions, where one motor is kept in a healthy state, and the other is subjected to a broken rotor bar (BRB) fault. The induced electromotive force of the triaxial coils, recorded over ten days with 100 measurements per day, is statistically analyzed. Normality tests and graphical interpretation methods are used to evaluate the data distribution. Parametric and non-parametric approaches are used to analyze the data. Both approaches show that the measurement method is valid and consistent over time and statistically distinguishes healthy motors from those with BRB defects when a reference or threshold value is specified. While the comparison between healthy motors shows a discrepancy, the quantitative analysis shows a smaller estimated difference in mean values between healthy motors than comparing healthy and BRB motors.