Development of asynchronous motor bearing fault diagnosis method using TDA and FFNN

Abstract

Asynchronous motors (AM) are life line of any process industry. Malfunctioning of AM at any stage of process leads the cost of finish product and decrease the efficiency of plant. Hence detection and diagnosis of AM failure at early stage is essential for timely maintenance and enhance the overall efficiency of the plant. The work present in this paper focuses on the bearing faults of AM. For this purpose experimental setup is developed in laboratory and results are based on experimental study carried out in laboratory by analysing AM generated vibration signals using time domain analysis (TDA) and feed forward neural network (FFNN).