Electrical fault diagnosis in induction motors using local extremes analysis

Abstract

Purpose – The purpose of this paper is to introduce an algorithm based only on local extreme analysis of a time sequence to further the detection and diagnosis of inter-turn short circuits and unbalanced voltage supply using vibration signals. Design/methodology/approach – The upper and lower extreme envelopes from a modulated and oscillatory time sequence present a particular characteristic being of, theoretically, symmetrical versions with regard to amplitude reflection around the time axis. Thus, one may say that they carry the same characteristics in terms of waveforms and, consequently, frequency content. These envelopes can easily be built by an interpolation process of the local extremes, maximums and minimums, from the original time sequence. Similar to modulator signals, they contain more detailed and useful information about the required electrical fault frequencies. Findings – Results show the efficiency of the proposed algorithm and its relevance to detecting and diagnosing faults in induction motors with the advantage of being a technique that is easy to implement in any computational code. Practical implications – A laboratory investigation carried out through an experimental setup for the study of faults, mainly related to the stator winding inter-turn short circuit and voltage phase unbalance, is presented. Originality/value – The main contribution of the work is the presentation of an alternative tool to demodulate signals which may be used in real applications like the detection of faults in three-phase induction machines.