To Develop a Deep Learning Based Algorithm to Detect & Classify Industrial Motor Faults (Condition Monitoring of Induction Motors)

Abstract

This thesis addresses the critical issue of condition monitoring for induction motors, a cornerstone in various industrial applications. Induction motors play a pivotal role in powering machinery and systems and their optimal performance is imperative for overall operational efficiency. Faults and failures of induction motors can lead to excessive downtimes. This motivates the examination of condition monitoring which is a technology that aims to find faults at the beginning of the fault, thanks to the data they collect by monitoring electric motors and rotating equipment, it detects the unexpected faults of a critical system. The study begins with a comprehensive review of existing condition monitoring methodologies, highlighting their strengths and limitations. It then introduces an innovative approach that uses vibration analysis to provide a holistic assessment of the motor's health. Machine learning algorithms are employed to process and analyze the data, enhancing the system's ability to detect subtle anomalies and predict potential faults. In conclusion, this thesis contributes to the field of condition monitoring by presenting a comprehensive and integrated approach for assessing the health of induction motors. The findings of this research have significant implications for industries relying on induction motors, offering a pathway to improved operational efficiency and reduced maintenance cost