کنفرانس بین المللی ماشین و محرکه های الکتریکی

صفحه اصلی / The 5th International Conference on Electrical Machines and Drives

Development of an ANFIS-Based Model for Condition Monitoring and Fault Diagnosis of Squirrel-Cage Induction Motors

نویسندگان :

Seyed Hamid Rafiei¹ Mansour Ojaghi²

1- zerc 2- Department of Electrical Engineering, University of Zanjan

کلمات کلیدی :

fault diagnosis،induction motor،condition monitoring،ANFIS

چکیده :

This paper proposes an Adaptive Neuro-Fuzzy Inference System (ANFIS)-based framework for condition monitoring and fault diagnosis of squirrel-cage induction motors. High-fidelity simulations were performed using Ansys Electronic Desktop to investigate the effects of three fault types: mechanical misalignment, broken rotor bars, and stator winding short circuits. Fault-specific features were extracted from the simulated motor current signals and employed to train the ANFIS model, which combines the interpretability of fuzzy logic with the adaptive learning capabilities of neural networks. The proposed approach effectively captures the complex nonlinear relationships between input features and motor operating conditions, achieving precise classification of different fault types. Validation results demonstrate 100% accuracy in both training and testing phases, highlighting the model’s strong generalization performance and robustness. This methodology minimizes reliance on costly experimental testing, enabling continuous online monitoring and predictive maintenance in industrial environments. Comparative evaluation with SVM, ANN, and CNN shows that ANFIS outperforms alternative methods in both accuracy and generalization, while maintaining moderate computational cost. Overall, the ANFIS-based model offers a reliable, interpretable, and efficient solution for enhancing the operational reliability and reducing maintenance costs of induction motor-driven systems.