In view of the technical problems in the process of rolling bearing fault diagnosis such as sample processing and identification of faults

a fault diagnosis classification method based on Morlet wavelets and classification and regression tree (CART) was proposed. Firstly

the Morlet wavelet analysis method and moving window method were used to process samples of the measured vibration signal of bearing. Secondly

the variational modal decomposition and feature extraction were performed on the extracted short samples to complete the construction of the training and test sets. Then

the training set was used to train the CART decision tree classification model

while random search and K-fold cross-validation were introduced to obtain the ideal classification model of bearing fault by optimizing the key parameters of the model. The test set validation results show that the method not only achieves effective diagnosis of various bearing faults and performs well in test sets with noise

but also significantly reduces the data length and sampling time of individual samples.