Fault Diagnosis in the Brushless Direct Current Drive Using Hybrid Machine Learning Models

The brushless direct current (BLDC) motor drive is gaining popularity due to its excellent controllability and high efficiency. This paper introduces a fault diagnosis method for open circuit (OC) short (SC) BLDC drives using hybrid classifier with optimization. Features such as current, voltage, speed, torque are considered the training data. features extracted by discrete wavelet transform (DWT) then employed train classifiers distinguish between types values of response parameters support vector machine Naive Bayes (SVM-NB). To further improve performance system, chaotic particle swarm optimization (CPSO) algorithms teaching-learning-based (TLBO) used. capable detecting recognizing type faults in motor. developed approach implemented on MATLAB/SIMULINK OC, SC, no-fault conditions. These provide better compared existing approaches respect sensitivity, accuracy, specificity. improved model achieves about 98.8% accuracy.