基于深度学习的采煤机截割部齿轮故障预测.

Aiming at the problem that the gear failure of the cutting part of the shearer reduced the working efficiency of coal mining production and brought hidden dangers to production safety, taking the MG1000/2500-WD shearer as the research object, the gears of the cutting part of the shearer were analyzed, and the failure mechanism research, simulation analysis and experimental test research were carried out. Firstly, the overall structure of shearer was analyzed, and the common causes and mechanism of gear failure of shearer cutting part were analyzed. Then, the expression formulas of convolution layer, pooling layer and full connection layer of convolution neural network were summarized. The gear fault model of cutting section was constructed based on deep convolution neural network (D-CNN), and the algorithm flow of the model was studied. Finally, by selecting the training data set, the model was trained, and the gear faults of the cutting part were predicted and classified; combining with the performance evaluation index, the experimental results of different models were compared. The experimental results show that the prediction model based on deep learning method can effectively predict the gear fault of shearer cutting part. The gear fault recognition rate is about 98.71%. Under the same conditions, the D-CNN model achieves an accuracy rate of 98.78% and a recall rate of 98.88% for the normal state and fault state of the gear. The model has a more accurate identification rate of gear faults. It has high fault prediction and classification performance. [ABSTRACT FROM AUTHOR]