A Reliable Neural Network Approach for Monitoring and Diagnosing Bearing Faults in Induction Motors

This study aims to develop a comprehensive approach for monitoring and diagnosing faults in induction motors, with a specific focus on bearing faults, which are the most prevalent type of fault in these motors. A monitoring system based on artificial intelligence techniques and a multilayer perceptron neural network is utilized to predict these faults. The artificial network is trained using the Levenberg-Marquardt algorithm, optimizing its performance as a diag nostic tool for induction machines. Temporal vibration indicators obtained from vibration analysis serve as inputs to the network, enabling the identification of fault presence or absence. The obtained results demonstrate the significant influence of training and test parameters and data on the neural network's performance. This research provides valuable insights into the application of artificial neural networks in fault diagnosis for induction motors, offering a reliable approach for proactive maintenance and enhancing the overall reliability of industrial motor systems.