Design Of Smart Electric Energy Meters Error Monitoring System Based On Big Data

Since 2009,State Grid Corporation of China has gradually promoted the pilot installation and measurement of smart energy meters.Up to now,Liaoning Electric Power Co.,Ltd.has more than 25 million electric energy meters connected to the grid.According to the regulations for the measurement and verification of electric energy meters,electric energy meters that have been installed for more than 8 years have gradually reached the legal service life.After the expiration,they need to be dismantled,and new meters must be purchased and replaced.The rotation work requires a lot of personnel and resource support.Relying on the big data analysis technology,this paper studies and constructs the operation error calculation and abnormal error diagnosis model,and realizes the online monitoring and remote control of the field error of the electric energy in operation.Specifically,the following research work is carried out:(1)Develop a strategy for the storage and processing of massive data of smart energy meters.Based on the massive data characteristics of electric power big data,ES cluster is clearly adopted as the real-time computing engine technology by comparing different big data processing schemes,and the fast distributed computing scheme is realized by optimizing the index and sharding strategy to realize the big data storage function.The basic strategy for data extraction,caching,and storage was formulated,and the basic data was initially cleaned and supplemented to lay a solid foundation for subsequent data applications.(2)Establish a calculation model of the electric energy meter error based on the law of energy conservation.In terms of model construction,the platform model is first constructed by the principle of energy conservation,the quasi-Newton method is introduced for preprocessing of iterative convergence calculation,and the recursive least squares method is used to complete the error fitting calculation.By analyzing the load characteristics of unconventional platform areas such as newly-built station areas,large-user station areas,and light-load station areas,and establishing mathematical models to optimize the basic model,solve the problems of underdetermined equations and insufficient calculation accuracy,and realize the inability to operate from the station area.The transformation from low-confidence to high-confidence calculation can improve the calculation accuracy of the model.(3)Based on the field feedback results and the operation principle of the electric energy meter,five types of abnormal measurement errors are determined,the abnormal causes and corresponding data characteristics are clarified through abnormal data analysis,and the KL divergence method is used to learn and extract various abnormal characteristic information from historical data.On the basis of data feature confirmation,XGBoost decision tree algorithm is used to construct various abnormal diagnosis models,and grid search and four-fold cross-validation are used to achieve continuous optimization of model parameters.The accuracy of the model is confirmed through the calculation results of abnormal instances on site,and the cause diagnosis of abnormal error values of measuring equipment is realized.(4)Based on the energy meter error calculation and measurement abnormality diagnosis model,a smart energy meter error monitoring system is designed and implemented.Taking the electricity and load data in the electricity consumption information collection system and the archive data in the marketing business application system as the input data of the model,through data calculation and model training,the error value of the on-site electric energy meter and the measurement abnormality diagnosis of the out-of-tolerance electric energy meter are obtained.,according to the dimensions of bidding batches,equipment manufacturers,etc.,to realize the classification and statistics of abnormal electric energy meters,and to provide work order circulation and on-site evaluation feedback solutions for abnormal resolution.The system provides remote online means for on-site error monitoring,liberates part of manpower to operate,and proposes an effective solution for improving the operating life of the electric energy meter.