AN EXPERIMENTAL EVALUATION OF INTELLIGENT FAULT DETECTION AND CLASSIFICATION FOR INDUCTION MOTORS UTILIZING MACHINE LEARNING APPROACHES

Abstract

Maintenance planning is critical for efficient operations of manufacturing systems. While unnecessary maintenance causes waste of money and time, skipping necessary maintenance can also cause unexpected down times in production. Predictive maintenance activities which focus on both detection and classification of equipment faults at an early stage are classified under Condition-Based Maintenance. On the other hand, forecasting remaining useful life of equipment is classified under Prognostics. In our study, fault detection and diagnosis of induction motors which are widely used in factories for different purposes are targeted. Triaxial vibration data collected from two similar induction motors under different operating conditions are examined for potential failure scenarios. Various features of vibration data are extracted, scaled and labeled with operational status information. The obtained dataset is analyzed with six different machine learning algorithms. Model performances are examined and compared against each other. Our experimental results show that the abnormal operating conditions of induction motors can be successfully detected utilizing machine learning algorithms.

Keywords

• predictive maintenance
• machine learning
• vibration analysis