Abstract

Three-phase induction motors (IMs) are fundamental components in industry, and like all mechanical devices are liable to failure that can lead to the shutting down of an industrial process. Rotor bar breakage is one cause for failure, and early detection of this abnormality would help to avoid costly breakdowns. However, the detection of broken bars in squirrel cage induction motor has been an important but difficult challenge. In this paper, a novel approach, based on chaos theory, is presented for detecting broken rotor bars. A chaotic oscillator called the Duffing oscillator is used for fault diagnosis. The presence of a fault is detected from a phase transition from chaotic motion to periodic motion in the Duffing oscillator phase plane. For automatic detection of this phase transition in Duffing oscillator, phase trajectory autocorrelation is introduced and used in this paper. Binary outputs along with simple calculation and high precision are the main advantages of this method. The combination of experimental and simulation results demonstrates the effectiveness of the method and the suitability of Duffing oscillators  for diagnosing faults due to broken rotor bars.