Reliable Detection of Rotor Bars Breakage in Induction Motors via MUSIC and ZSC

Induction motors are used in a variety of industrial applications where frequent startup cycles are required. In those cases, it is necessary to apply sophisticated signal processing analysis methods in order to reliably follow the time evolution of fault-related harmonics in the signal. In this paper, the zero-sequence current (ZSC) is analyzed using the high-resolution spectral method of multiple signal classification. The analysis of the ZSC signal has proved to have several advantages over the analysis of a single-phase current waveform. The method is validated through simulation and experimental results. The simulations are carried out for a 1.1-MW and a 4-kW induction motors under finite element analysis. Experimentation is performed on a healthy motor, a motor with one broken rotor bar, and a motor with two broken rotor bars. The analysis results are satisfactory since the proposed methodology reliably detects the broken rotor bar fault and its severity, both during transient and steady-state operation of the induction motor.