Research And Application Of Smart Grid Electricity Demand Forecasting Technology For Smart City

The conception of smart city has been developing rapidly from the aspects of conceptual model to the planning and construction level since 2009,and has set off a worldwide trend of smart city planning and construction.Smart city's main task is inter-departmental big data operation and management.Aiming at urban integration,it establishes a unified system of data collection and sharing across departments to provide convenient data services for intelligent application in all areas.As an important infrastructure of smart city development,the smart grid also plays a key role in smart city construction,which in turn meets the objective need of the smart city development.Electricity demand forecasting is the most important part in power industry activities and the significant support for power enterprises'survival and development.Firstly,the article is different from the traditional big data forecasting model of the electricity industry.According to the characteristics of the smart city big data,the author put forward a new thought of electricity demand forecasting.The meteorological information,geographic information,population information,enterprise information,economic information and other external data should be included into the database.And finally it forms a more intelligent trend analysis,providing decision support for decision makers.Secondly,this thesis focuses on two key business issues of smart grid.On the one hand,build electricity customer segmentation model.On the basis of the smart city's basic database,the author tries to achieve scientific customer cognition by identifying the ontology characteristics and behavior characteristics of different customer groups,and put insurance information,per capital production value and legal entities' information into consideration,and finally use the K-means algorithm to implement data mining.Optimize the electricity customer information to improve its segmentation precision,and finally provide comprehensive reference for electric power department's risk management,personalized marketing and development plans.On the other hand,on the basis of the electricity customer segmentation and the smart city basic database,consider other big data information categories,and put the population information,enterprise information,and macroeconomic information into a unified analysis.The thesis use BP neural network algorithm to build electricity prediction model,and evaluate electricity market from multidimensional aspects through exploring the relationships between data,and finally make exact and scientific prediction on electricity demand trends,which cannot be achieved through traditional prediction methods.At last,on the basis of some city's smart city public information service platform,the author analyzes its function and data processing of each part.The research result is applied to retrieval of power information system.The author design the algorithm parameter adjustment function for prediction model,power demand forecasting function,and the electricity customer segmentation function in detail.This system is based on B/S architecture,and is encoded by Java,Flex and so on.Each of the function module operates normally.