A Multi-sensor Fusion Scheme for Broken Rotor Bar and Air-gap Eccentricity Detection of Induction Machines

A new multi-sensor fusion scheme for induction motor broken rotor bar fault and dynamic air gap eccentricity fault detection is proposed in this paper to refine the sensor selecting process and improve the correlation of data. Instead of combining the signatures for the same fault, the proposed scheme combines the characteristic frequencies from different sensor related to the same abnormal phenomenon such rotor asymmetry and speed ripples. An experimental validation is performed on a 7.5 horsepower induction motor. The data for unbalanced load condition is also gathered, and it can be separated from similar condition. The experimental results validated the new fusion scheme and proved that with the proposed scheme, higher sensitivity and specificity can be achieved.