Abstract:

The condition monitoring of induction motors is critical for ensuring their reliable operation and minimizing downtime in industrial applications. Traditional monitoring methods often fall short in providing timely and actionable insights, leading to increased maintenance costs and potential system failures. Voltage, current, temperature, and vibration are some of the induction motor metrics used to track the health of IMs in this study. In addition, the analyses are conducted using two distinct Machine Learning (ML) techniques, namely, k-means and extreme learning machine. In addition, we compare the two ML algorithms' performance and use the one that performed the best as a classifier in our newly created intelligent system that operates in real-time. The results show that compared to the k-means classifier, the Extreme Learning Machine (ELM) classifier performs better.

Description:

.