Research And Implementation Of Monitoring And Management Technology For Electric Heating Equipment Based On Edge Computing

With the development of science and technology,the era of interconnection of all things is getting closer and closer to us.In the field of smart home,great progress has been made.However,with the rapid increase in the number of network edge devices,these devices will generate huge amounts of data.The large data computing method with cloud computing as the core can no longer meet the fast-growing demand for computing data.Therefore,the edge-based large data processing method based on the edge computing model has been paid more and more attention.In order to ensure the smooth operation of household equipment,it is necessary to monitor and manage the equipment in real time.Traditional cloud-based smart home equipment monitoring needs to transmit sensing data to the cloud through network,which is affected by network transmission,and can not guarantee the real-time monitoring of home equipment.Therefore,this paper proposes a monitoring and management technology of electric heating equipment based on edge computing,which combines cloud and edge nodes to monitor electric heating equipment.It can reduce response time,improve data transmission efficiency and realize real-time monitoring and management.The main work of this paper includes:Firstly,by analyzing the requirement of monitoring and management of electric heating equipment,the overall structure of monitoring and management system of electric heating equipment based on edge computing is expounded.The architecture is mainly divided into three modules:cloud monitoring management module,edge device monitoring management module and user mobile module.The edge is responsible for receiving sensor data generated by heating equipment and calculating them in real time according to monitoring requirements.The results are sent to the cloud for persistent storage.The cloud is responsible for computing tasks with large amount of computation and low real-time requirements,and provides external data source data for edge nodes.Therefore,based on edge cloud collaborative monitoring,real-time monitoring and management of equipment can be ensured.The edge end will return the monitoring information to the mobile end in real time,and the user will know the equipment monitoring situation through the mobile end in time.Secondly,based on the analysis and research of equipment monitoring technology,combined with the practical application requirements of heat storage electric heating equipment,this paper proposes a monitoring and management method of electric heating equipment based on edge calculation,which is divided into abnormal detection of electric heating and prediction of heating.The abnormal detection of electric heating equipment is to monitor the real-time sensing data,to judge and analyze the abnormal faults,and to remind users to eliminate the faults in time.Heating prediction of electric heating equipment is based on users’ preferences for heating in different weather conditions.LSTM recursive neural network benchmark prediction model is built to predict the heating storage time required by users for the next day.Reasonable heat storage and heating planning can be realized to meet usersl’heating needs and reduce energy consumption at the same time.Thirdly,the monitoring and management system of electric heating equipment is designed and implemented,including the monitoring and management system at the edge,which provides two functions:anomaly detection and heating prediction;cloud monitoring platform,which provides Web interface to view the operation of electric heating equipment;and mobile APP,which provides real-time monitoring function of electric heating equipment.Finally,the experiment validates that the response time of the anomaly detection algorithm under two computing modes is compared in experiment 1.The results show that the response time of the anomaly detection algorithm is faster in the way of edge cloud collaboration.The second experiment is the evaluation experiment of the accuracy of the heating prediction.The results show that the forecasting method proposed in this paper can reasonably plan the user’s heat storage time.Compared with the traditional BP neural network,it can forecast the user’s heating data more accurately,meet the user’s daily heating needs,while effectively avoiding the energy consumption problems caused by excessive heat storage.The third experiment is the comparison of training time between cloud and edge using warm prediction algorithm.The results show that the training time in cloud is shorter and the method of edge cloud collaboration is better.