Time-domain estimation of sub-harmonic sinusoidal disturbance in sinusoidal signal with applications in induction motor diagnostics

One of the most important signals that we come across everyday in many fields is the sinusoidal signal. However, in many real-life applications, the sinusoidal signals of interest are contaminated with disturbance signals. One of the interesting cases is when the disturbance signal is also sinusoidal, but not necessarily of the same frequency and amplitude as the original signal. Possible approaches to deal with the problem involve frequency analysis of the signal, adaptive state estimation, etc. In this paper, a time-domain approach is proposed for sub-harmonic classes of sinusoidal disturbance, which is based on the detection of the envelope of the measured signal. This would require significantly reduced computation, for example, without needing computationally expensive discrete Fourier transform (DFT), adaptive filters, etc. The theoretical analysis of the proposed method is substantiated by application results, and possible use-case of online diagnostics of broken rotor bar for induction motor.