RESEARCH ON FAULT FEATURE EXTRACTION OF ROLLING BEARING BASED ON IMPROVED CEEMDAN

Abstract

As a core component of machinery, rolling bearings play an important role in modern industry.In order to avoid unpredictable accidents and undesired downtime costs, fault diagnosis techniques for rolling bearings are widely used in the industrial field.Although there are many troubleshooting methods, such as empirical mode decomposition (EMD), collective empirical mode decomposition (EEMD) and wavelet transform, their disadvantages are very obvious.For example, the eigenmode function (IMF) generated by EEMD always contains residual noise.Furthermore, adding different types of Gaussian noise to the signal will result in a different number of IMFs, which makes the average difficult.In order to eliminate these shortcomings, this paper proposes an improved combination of wavelet transform and full set empirical mode decomposition and adaptive noise (CEEMDAN).Simulation experiments were used to illustrate the effectiveness of CEEMDAN.To further illustrate the proposed method, the improved CEEMDAN is applied to fault diagnosis motor rolling bearings.In addition, two indicators, signal-to-noise ratio (SNR) and root mean square error (RMSE), are established to evaluate the fault diagnosis capability of the proposed method.The analysis results show that the method can improve the fault diagnosis ability of rolling bearings.