Fault Diagnosis of Multistage Gearbox Based on Adaboost Ensemble Learning

In this paper, by building a multi-stage gearbox fault simulation test bench, the operation status of the gearbox in six different fault states is simulated and the gear vibration signals are collected. The signal is decomposed into different IMF components from high to low by Ceemdan and fault sensitive features are extracted. By constructing the SVM-Adaboost ensemble learning model, the Ceemdan time-frequency energy feature is used as the model input for fault diagnosis of multi-stage gearboxes.