Detection of Stator Interturn Fault in Induction Motors Using SNR Concept

The stator interturn fault (SITF) in induction motors (IMs) is recognized as one of the most common electrical failures, arising from the various stresses encountered during operation. Consequently, monitoring the health of IMs, as the most widely used electric machines, is of paramount importance in both industrial and transportation applications. This study introduces an innovative algorithm for detecting SITF. The algorithm functions by subtracting two resized data windows, using the corresponding lag of the maximum sample cross correlation value to extract SITF characteristics. The signal-to-noise ratio (SNR), defined as the ratio of the power of the low-frequency signal to that of the fault signal, serves as the evaluation criterion. Notably, this criterion demonstrates robustness against nonfault conditions, such as significant mechanical load variations and voltage quality issues, including sags, swells, and imbalances. A series of experimental scenarios are conducted to assess the efficacy of the proposed SNR-based algorithm, confirming its high accuracy, manageable computational burden, and excellent response time.