Fault Diagnosis of Rotating Machinery Based on Multiscale Entropy

Abstract

The accuracy of fault diagnosis is directly determined by the accuracy of fault information classification of rotating machinery. Based on the analysis of the basic theories of sample entropy and multi-scale entropy, and through the comparative analysis of sample entropy and multi-scale entropy on the original experimental data, it can be seen that multi-scale entropy is able to classify fault information more effectively. The optimal scale was determined through Matlab analysis. Under this scale, the fault differentiation capability under different similar tolerance was simulated, and the optimal scale and similar show that the multi-scale entropy is effective in differentiating various faults under the same scale and similarity tolerance.