Research And Implementation Of Deep Learning In Vibration Failure Prediction Technology Of Rotating Machinery

With the advancement of modern industry and the improvement of automation,the connections between various equipment in rotating machinery systems are becoming closer.Once a large machine fails,it will adversely affect production efficiency and production quality,and even endanger the life safety of field workers.Predicting the failure of rotating machinery can provide early warning before a failure occurs,so that workers can take timely action to reduce losses.Experts at home and abroad have conducted extensive research on the subject of fault prediction,and the development of big data and artificial intelligence has provided a new development direction for vibration fault prediction of rotating machinery.Aiming at the above background and research status,this paper conducts research on vibration fault prediction technology of rotating machinery based on deep learning.First,the basic theory of deep learning is introduced,which provides a theoretical basis for subsequent research.Most existing fault prediction models only consider the forward information of time series data and do not consider the backward information of time series data.Therefore,this paper proposes a vibration failure prediction model for rotating machinery based on bidirectional long short-term memory network and attention mechanism,which fully considers the information of the time series data in the forward and backward time directions,and highlights the data series that have a key influence on the prediction result,improving the accuracy of the prediction of the vibration data of rotating machinery.The purpose of prediction is to diagnose the fault.The predicted result is sent to the fault classifier for judgment.If the judgment is fault data,an alarm message is sent.In this paper,the proposed model is trained and verified through experimental bench data and field data.At the same time,by comparing with the other two prediction models,it is proved that the model proposed in this paper can effectively improve the prediction accuracy,and has higher validity in the prediction of vibration failure of rotating machinery.