Diagnosis of induction motors fed by supplies with high harmonic content using motor current signature analysis

Abstract

Electrical monitoring is currently considered as a standard in preventive maintenance and is the reference method for the diagnosis of medium-large machines, standing out the technique known as motor current signature analysis. Numerous successful line-driven motor fault detection methods are reported in the literature. However, inverter-fed motor systems and utility driven motors with high harmonic content still require more attention. In these cases, fault detection is more difficult due to remarkable noise on the line current, transient operating conditions, closed-loop controller bandwidths and, above all, the presence of many time harmonics introduced by the supply. This harmonic content, which is normally known as a major side effect, gives rise to extra signatures and turns out to be useful in distinguishing faulty current spectrum patterns from healthy ones.