Design Of Residential Electricity Forecasting And Visualization Platform Based On Power Big Data

The era of big data has come quietly,and various industries at home and abroad are actively developing big data technologies to create value.With the implementation of the energy Internet and the new power reform,the massive historical data of the traditional power industry is also in urgent need of big data to bring about technological changes.How to use this massive historical data and discover important information from it has become a cutting-edge problem in the current power system.Based on this idea,this paper studies the development process of big data at home and abroad,and how various industries use big data to create value.And this combination of historical data,analysis of user characteristics,and the creation of useful value ideas in the power industry.Through clustering,forecasting and visual display,the staff is provided with a more convenient data display platform for staff to view,count and mine data.On this basis,improve the management and monitoring of user power information.Load forecasting is related to electric energy production.It has always been an important basis for power dispatching: the law of electric energy demand is directly related to the arrangement of power generation plan.The optimization of power generation plan is the most basic means of economic operation of the entire power system.This paper relies on big data technology,combined with meteorological data,to analyze the historical electricity consumption of some users in Hanzhong City.After preprocessing the data and solving the bad fields such as missing values and outliers,the K-means algorithm is used to cluster the users and analyze the different power usage of various users.With the help of clustering results,the neural network algorithm using two different structures,RBF and MLP,is used to predict the user's electricity consumption,and the prediction results of different neural network structures are compared to select higher accuracy and lower average relative error.The prediction results of the MLP neural network serve as a new data source.On this basis,using the latitude and longitude information,make a visual map through Tableau,position the user on the map,and create an array visualization worksheet to display the user's electricity consumption information.Finally,through the calculation field,information such as user location,historical power,user clustering and power forecasting are displayed together and presented to the staff in a visual interactive interface.Linking the map to the statistical worksheet,using the superiority of the interactive interface,brings more convenience to the staff,and helps the lean management of the power information.