The underground shearer structure is complex,the working environment is poor,prone to all kinds of faults,and the disposal is difficult.According to the gear fault diagnosis requirements of MG1000/2500-WD shearer,combined with the fault feature extraction principle of CNN(convolution neural network),Fenxi Mining Group Equipment Repair Factory designed a DCNN(deep convolution neural network)fault prediction model,and implemented the data training and model testing.The results show that the fault screening accuracy of the prediction model reached 98.17%,the accuracy of fault training value and standard value reached 99.13%.
维普
万方数据