Overload Monitoring System Using Sound Analysis for Electrical Machines

Abstract

Overloading is one of the faults that occur very often in the operation of electrical machines. Therefore, a continuous monitoring and diagnosis for this is necessary in safety-critical applications. This paper presents a sound analysis system used for detecting and classifying induction motor and power transformer overload levels with a microphone. Three acoustic features and six classification models are evaluated. The obtained results show that this is a promising way to monitor electrical machines overload.