Influence of the Flux Sensor Position for Rotor Fault Detection in WRIM: A Power Spectral Entropy Analysis.

Over recent years, the analysis of stray-flux data has revealed itself as a powerful technique for the diagnosis of several faults in induction motors. The combined application of steady-state and transient analysis tools to stray-flux data has proven to be effective not only to detect but also to discriminate between different types of faults raising in these machines. In spite of its proven potential, one of the pending concerns relies on the influence of the position of the flux sensor on the obtained results. In this context, different authors have suggested alternative positions to capture the stray-flux data. This work deepens in the analysis of the influence on the flux sensor when diagnosing rotor asymmetries in Wound Rotor Induction Motors (WRIM). The considered diagnosis method relies on the analysis of stray-flux data under starting, by applying the power spectrum entropy. Moreover, the Pearson correlation coefficient and the histogram of the entropy of the power spectrum are used to corroborate the probability distribution of the spectrum in each relative position. The results are useful to indicate the sensitivity of each sensor position for the detection of the considered failure.