Abstract

Three-phase induction motors are essential in industrial applications due to their durability, efficiency, and reliability. However, faults in rotor bars can degrade their performance, leading to operational inefficiencies, unplanned downtime, and increased maintenance costs. This study presents a comprehensive approach to modelling and simulating three-phase induction motors with rotor bars, utilizing machine learning techniques to improve fault detection, diagnostics, and predictive maintenance. A mathematical model is developed to represent the motor's electromagnetic and mechanical dynamics, including rotor bar characteristics. Simulations are conducted under varying operational conditions, generating data that includes both healthy and faulty motor states. This data is used to train machine learning models such as Support Vector Machines, Random Forest, and Neural Networks, focusing on identifying fault patterns, classifying motor conditions, and predicting the remaining lifespan of rotor bars. The performance of the machine learning models is evaluated using metrics like accuracy, precision, and computational efficiency. Results demonstrate the potential of these approaches to enable accurate fault diagnosis and support realtime predictive maintenance strategies. This research contributes to the advancement of intelligent motor diagnostics by integrating advanced simulation techniques with machine learning. The findings offer practical benefits for industries aiming to reduce downtime, optimize energy efficiency, and lower maintenance costs.