Research On The State Data Monitoring And Abnormality Prediction System Of Air Compressor Based On Edge Cloud Collaboration

As a more commonly used industrial equipment,air compressor(air compressor)has become an indispensable equipment to provide power for intelligent manufacturing equipment,in petroleum,chemical,medical,natural gas transportation,automobile,mining,aerospace and been widely used in other industrial fields.However,the air compressor structure is relatively complex and the air compressor has a long running time,which easily leads to the abnormal operation state of the air compressor,and then affects the stability and reliability of the whole equipment or system.Therefore,it is of great significance to carry out the real-time monitoring and abnormality prediction system research to ensure the stable and reliable operation of intelligent manufacturing equipment.The research work on the air compressor state data monitoring and exception prediction system based on the edge cloud collaboration technology mainly includes:(1)The background and practical significance of the research on air compressor status data monitoring and anomaly prediction system based on edge cloud collaboration are expounded in detail,and the development and application status of industrial equipment operation status data monitoring and anomaly prediction technology,edge computing and edge cloud collaboration technology are summarized.Through the fault analysis of air compressor,the main forms of abnormal state operation data are defined,which provides theoretical basis for the real-time monitoring and abnormal prediction algorithm of subsequent state operation data of air compressor.(2)According to the problem of misdetermining the threshold of the abnormal data,a real-time monitoring algorithm of the operation state data of the air compressor based on BGM-MD is proposed.By collecting the current signal data and vibration signal data of the air compressor,the variational Bayes is applied to the Gaussian hybrid model to solve the problem of hyperparameters;combined with the Mahalanobis distance algorithm to judge the data anomaly,and the high-dimensional anomaly data is converted into one-dimensional data to judge and determine the threshold,so as to reduce the misjudgment of the running state data of the air compressor.The experimental results show that the proposed method determines the threshold of abnormal data more simply and can effectively reduce the miscalculation rate of normal data.(3)For the prediction problem of abnormal operating state data of air compressor,an abnormal prediction algorithm of air compressor state data based on LSTM network model is proposed.The model parameters are selected through the evaluation index,the LSTM network model is constructed,and the model parameters are adjusted and optimized according to the variable control principle.The BGM-MD algorithm and the LSTM network model to realize the function of device exception prediction.The experimental results show that the prediction data can better match the real data,and can effectively realize the abnormal prediction of the air compressor running state data.(4)Based on the air compressor state data monitoring and abnormal prediction system response real-time problem,this paper adopts the edge cloud collaborative technology,through the edge of data acquisition,saving and preliminary processing calculation,complete the depth in cloud computing end complex tasks,the edge cloud collaborative way can greatly reduce the data,so as to reduce the link transmission pressure and server task processing pressure.(5)Taking the air compressor applied in the high and low temperature test box in the project as the research object,the operation status data monitoring and abnormal prediction system of the air compressor based on edge cloud collaboration is designed.Using Bootstrap and Django framework,combined with data acquisition,data storage,data processing,function migration,network communication,data visualization technology to develop the system of front,design the state of the air compressor data monitoring and abnormal prediction system,implements the status of test box air compressor monitoring and abnormal early warning,ensure the stable and reliable operation of the equipment.