Rolling Bearing Fault Diagnosis Based on Fractal Dimension

摘要

Fractal, as a new technique of signal processing, is suitable for analyzing the non-linear fault signals of rotating machine. By researching the characteristic of non-linear vibration signals of rolling bearings, a study of box dimension in analyzing the vibration signals and diagnosing the fault pattern of rolling bearings is proposed. Box dimension algorithm is presented in details and quantificational calculating of non-linear vibration signals generated by bearing system is also discussed. Experimental results show that kinematics mechanisms of rolling bearing result in the different working state, so the box dimensions are different evidently. The application of box dimension in monitoring working state is a new approach to promote the accuracy of rolling bearings.