Fault Mixing Prediction Analysis Model Based on Standard Gear Reduction Box
In the pharmaceutical production line,agitator is the main component of the main reactor,it has the largest volume and relatively complex structure,and tandem relationship with other equipments.If a fault occurs,on the one hand,it is neces-sary to stop the entire production line for fault diagnosis and maintenance,resulting in idle loss of unplanned shutdown of equipment,on the other hand,sudden unplanned shutdown also causes the chemical material reaction substances in the equip-ment to be wasted due to insufficient reaction.In order to solve this problem,a fault mixing prediction analysis model based on standard gear reduction box is proposed.It fuses the converted frequency domain characteristic data and amplitude data accord-ing to the time series by feature extraction and variational mode decomposition of vibration signals,and divides the data into three categories:fault data,equipment sick operation data and equipment health operation data.Features are extracted through the self-attention network layer.Experiments show that the model can accurately predict the failure of the equipment,and the accuracy rate in the test set reaches 83.61％,which verifies the effectiveness and superiority of the experiment.
operation and maintenance managementfault mixing predictionfeature extractionvariational mode decomposi-tionself-attention network
万方数据
维普
