Prediction of common faults of shearer

The gear fault prediction model based on deep learning is designed by taking the gear of shearer cutting section as the research object and using the convolution neural network method, and the fault prediction and classification recognition ability of D-CNN model is verified by experimental method.

The research process and results are as follows:

(1) Analyzed the research status of gear fault prediction and diagnosis of shearer cutting part at home and abroad, elaborated the overall structure of shearer, and summarized the common fault types, fault causes and mechanisms;

(2) The structure of convolutional neural network is analyzed, and each type of neuron is designed, thus the fault prediction model is constructed, and the algorithm flow of fault prediction is developed;

(3) The model can accurately predict the fault state of the gear by using data training set to train the gear parameters. The model has good fault classification accuracy and recall rate, and can effectively predict gear faults. Compared with the recognition rate and classification performance of DNN, DBN and SAE models, the other three models have made great mistakes in the status of gear fault recognition, while the D-CNN based model can accurately identify faults.

In the subsequent research, feature parameter extraction of temperature sensor acquisition information will be added, and multisensor fusion research of fault prediction will be concerned, so as to mine more effective parameter features and make fault prediction methods more efficient and accurate.

Scroll to Top