Realization of Machine Learning Algorithms for Diagnosing Winding Faults in Induction Motors

The reliability and efficiency of squirrel cage induction motors (SCIMs) are most important in industrial applications, where failures can lead to significant downtime and economic losses. Among the common faults, stator winding faults are particularly detrimental, leading to degraded motor performance and potentially catastrophic failures if undetected at the incipient stage. Various approaches aiding machine learning algorithms are adapted for diagnosing machine faults. This paper realizes different machine learning algorithms to diagnose the winding faults of SCIM. A comprehensive dataset of motor current signatures is trained and validated with several machine-learning models using Python in the Anaconda platform. A detailed comparative analysis of the model's performance is conducted based on metrics such as accuracy, precision, recall, F1 score, and computational efficiency. This research highlights the potential of integrating machine learning techniques into SCIM monitoring systems, offering a proactive approach to maintenance and reducing unscheduled downtimes. Future work will explore the deployment of these models in real-time monitoring systems and the expansion of the dataset to include various motor types and fault conditions.