Prediction of radial clearance based on bearing vibration using artificial neural network

Artificial neural networks are used for diagnosing of ball bearings condition. They are capable to recognize bearing fault from information extracted from signal of vibration. The aim of the work is to investigate the relationship between vibration velocity in the characteristic band frequencies and radial clearance. Changing the radial clearance during operation may indicate changes in the operation of the rotating system. Experimental research is based on deep groove ball bearings and measuring the vibration of the bearings. The vibration velocities for each bearing were measured under laboratory conditions, and the processed signals were used to train the artificial neural network. An artificial neural network trained in this way can predict the size of the radial clearance based on the vibration of the bearing.