Current signature analysis of induction machine rotor faults using the fast orthogonal search algorithm

Abstract

This paper presents a method of detecting rotor faults in induction motors using the fast orthogonal search (FOS). Proper online monitoring of motors is important to ensure safe operation, timely maintenance, and efficiency. One method to detect rotor faults is to monitor the stator current spectrum for fault harmonics. Conventional methods that use the FFT for spectral analysis are inadequate for motors that are prone to transient conditions. Changing operating conditions during the sampling time may cause the spectrum to become smudged, reducing the amplitude of the fault signatures causing them to become undetectable. Reducing the sampling time minimizes the effects of smudging, however a fine resolution must still be maintained to distinguish between harmonics. This is especially important when operating under light-load since the fault signatures will be very close to the fundamental frequency. Therefore a balance between resolution and sampling time must be achieved, which is difficult with the FFT. This paper will demonstrate through simulation that FOS is able to achieve eight times the resolution of the FFT for the same data length. Therefore FOS is a promising choice for rotor fault detection in motors where long periods of steady-state are unavailable, and for motors operating with less than full-load.

Oblique projection pre-processing and TLS application for diagnosing rotor bar defects by improving power spectrum estimation

“…These kinds of approaches were certainly first confidence approach giving reliable results. However, for broken rotor bars detection, an excellent accuracy location of the sidebands components is essential [7,28]. Then, overcoming the resolution insufficiency of periodogram-based approaches which we remind, is limited by the Fourier resolution, i.e.…”

“…Owing to its non-invasive or non-intrusive capability [24], stator current acquisition is increasingly put in place. For squirrel-cage rotor induction motors, electrical defects such as rotor bar defects could be seen inside the stator current signal by monitoring the frequencies around the supply frequency [1,7,28]. In such an environment, the driven frequency is undisputedly the most powerful frequency inside the stator current spectrum and it may mask the meaningful sidebands frequencies.…”