Application of Artificial Neural Network in the Investigation of Bearing Defects

Maintenance and design engineers have great concern for the functioning of rotating machineries due to the vibration phenomenon. Improper functioning in rotating machinery originates from the damage to rolling element bearings. The status of rolling element bearings require advanced technologies to monitor their health status efficiently and effectively. Avoiding vibration during machine running conditions is a complicated process. Vibration simulation should be carried out using suitable sensors/ transducers to recognize the level of damage on bearing during machine operating conditions. Various issues arising in rotating systems are interlinked with bearing faults. This paper presents an approach for fault diagnosis of bearings using neural networks and time/frequencydomain vibration analysis.