Stator winding fault prediction of induction motors using multiscale entropy and grey fuzzy optimization methods

• The prediction of stator winding faults using multiscale entropy is performed for the first time.
• Real-time vibration and current are used as diagnostics to identify faults.
• The system complexity associated with motors is investigated using multiscale entropy.
• GFRG is used to predict fault and also to suggest optimal settings for motor operation.
• The motor condition has a maximum contribution of 54.21%, as determined from the ANOVA analysis.

In the present work, stator winding fault prediction is studied using a multiscale entropy (MSE) algorithm combined with a grey-based fuzzy algorithm. Experiments were performed with a normal motor and a motor with faulty stator winding. Real time, motor current and vibration signals were acquired at different operating speeds and were used for the diagnosis of faults. The obtained signals were denoised by wavelet transform. Grey relational analysis (GRA) coupled with fuzzy logic was used to model the stator winding fault and to predict the optimal setting for running the induction motor within its parameters range. Analysis of variance (ANOVA) was performed to determine the effect of each individual parameter on the response. The results indicate that the proposed novel approach is very effective in predicting the stator winding fault. Furthermore, the best running parameters for the induction motor are also reported.