Gearbox bearing diagnosis based on Sanc and 1d-cnn

Combined with the signal separation ability of Sanc and the intelligent feature extraction and fault identification ability of 1d-cnn, this paper proposes a gearbox bearing fault diagnosis method based on the combination of Sanc and 1d-cnn.

The main steps include:

(1) the random signal including bearing features is extracted by Sanc filter.

(2) standardize the random signal to make each index in the same order of magnitude and dimension, so as to speed up the learning efficiency of the model.

(3) the standardized data are randomly scrambled and divided into training set, verification set and test set according to a certain ratio.

(4) 1d-cnn neural network model is constructed to initialize the model parameters to facilitate the updating of model parameters.

(5) start the first round of training. After all batches of samples have been trained, the validation set is used to preliminarily evaluate the model and start the next round of training. Repeat the process until all rounds of iteration are completed.

(6) after the training, the test set is used to evaluate the final training effect of the model.