Abstract—This paper presents research on the recognition of induction motor overload levels using sound analysis. Stable and durable operations of induction motors are very important in home appliances and industries. Overloading is one of faults that can shorten the operating life of these electromechanical machines. In our studies, five levels of overload status are classified using sounds collected by a single microphone. Three acoustic features and six classification models are evaluated. The accuracy rate of 94.85% shows that this is a promising way to classify and therefore to monitor induction motor overload. 

Index Terms—induction motor overload, classification, sound analysis, machine learning

Cite: Nguyen C. Phuong, "A Five-Class Recognizer for Sounds of Induction Motor Overload," International Journal of Electronics and Electrical Engineering, Vol. 9, No. 3, pp. 59-64, September 2021. doi: 10.18178/ijeee.9.3.59-64