Entropy Approach Using PCA for Sequential Accelerated Aging Processes in Electrical Motors

The aim of this paper is to monitor the bearing damage developing by vibration signals measured from induction motors subjected to bearing fluting aging. The vibration signals taken from the accelerometers were analyzed by principal component analysis (PCA) and computed their spectra to denote the progressing of the aging. Meanwhile, using the principal component (PC) results, the entropy values based on the PC scores were computed and aging process models were established by the exponential and polynomial approach. Hence, maximum entropy value was determined to indicate the critical aging level.