Application of Teager Energy for Broken Rotor bar Fault Detection Based on the Motor Current Signature Analysis

This paper investigates a different implementation to achieve Teager energy for broken rotor bars (BRB) fault detection of industrial motors (IMs) based on motor current signature analysis (MCSA). The Teager energy can be obtained by either Teager-Kaiser energy operator (TKEO) or frequency domain energy operator (FDEO). TKEO is carried out by calculating three adjacent samples in the time domain with high efficiency, but the causal processing needs to be applied to avoid the phase distortion and non-ideal filter characteristics to ensure the performance of TKEO. In addition, one of the essential elements of the instantaneous frequency (IF) obtained by TKEO is the TKEO of the derivation of the signal, which leads to the result lacks physical meanings. By contrast, FDEO is a new method which estimates the Teager energy of a signal in the frequency domain with non-causal processing. And Teager energy can be estimated by calculating the square envelope of the first derivation of the signal. Hence a BRB fault diagnosis method is proposed based on the FDEO of the motor current to remove the influence of the main supply frequency followed by spectral analysis for fault feature extraction. The proposed approach is validated in experimental research where healthy IM and IMs with one BRB and two BRBs are operated under different loads. It certificates the FDEO provides slightly obvious fault features of TKEO, especially when the IMs works under low load conditions.