Fault Diagnosis of Motor Based on VMD-Sample Entropy-Random Forest

Abstract

Traditional fault diagnosis of DC motors methods need to establish accurate mathematical models, effective state estimations, estimation of parameters and appropriate statistical decision-making methods. These preconditions make the traditional motor fault diagnosis have considerable limitations. To address this issue, a new mechanical fault diagnosis method is proposed. Firstly, the vibration signals of rotary bearing are collected by the designed acquisition system. Then, the variational mode decomposition (VMD) is adopted to decompose the signal into a series of intrinsic mode function, and the characteristics of the vibration signal are extracted by sample entropy. Finally, random forest uniting SPRINT algorithm is used to identify vibration signals of rotating machinery, which each branch tree is trained by applying different bootstrap sample sets. The results have shown that the proposed fault diagnosis method has good generalization performance and the recognition rate samples is more than 90%. Compared with traditional neural network, and it has no use for carrying out the verbose process of parameter optimization.