Efficient Real Time Stator Inter-Turn Short Circuit Fault Detection of Inverter-Fed Induction Motors: A Hybrid Approach Utilizing Wavelet Transform and Digital Signal Processing

In induction motors faults in the stator winding can significantly impact their performance and reliability. This research investigates the analysis of inter-turn short circuit faults by varying the switching frequency of the sine pulse-width modulation (PWM) inverter under different loading conditions. A faulty condition of the motor is modeled in MATLAB/Simulink blocks, and stator currents are simulated for different fault percentages. First, the wavelet transform is applied to decompose the stator current signals into different frequency sub bands. Next, statistical parameters are computed from the wavelet coefficients to quantify the characteristics of the stator current signals. These statistical parameters can include mean, standard deviation, variance, skewness, kurtosis, or other measures that capture the statistical properties of the wavelet coefficients. This Statical data is used to train the SVM model and find a optimal decision boundary to classify Healthy or Faulty Condition.