Load Characteristics Analysis And Forecasting Based On Intelligent Electricity Utilization Information Collection System

With the development of a strong smart grid,more advanced sensor measurement,communication network,automatic control,information technology,new materials and other technologies integrate into the grid,the traditional power grid become more intelligent and information.In the smart grid service system architecture,smart electricity information utilization system can collect and process the real-time data.Load characteristics and forecasting based on collected data can provide a more scientific basis for the planning of power grid dispatching department.So as to improve the proportion of energy consumption in the terminal energy consumption,to achieve the goal of peak shaving and valley filling,energy efficiency,energy saving and consumption reduction,and it is a great significance to the safety,efficiency and economic operation of power grid.Analysis of load characteristic is ahead of load regulation,which is the basis of load forecasting,and it is also an important content of power demand side management.This paper based on intelligent electricity utilization information collection system combining with the real-time data analysised the load data in different time scales including the load data of year,month,day.The effects of economic,time series,weather and other aspects on load characteristics a re described.Combined the charts show the correlation between the load and maximum temperature,minimum temperature,average temperature,relative humidity,average wind speed.It provides basis for load forecasting.Load forecasting is the scientific basis of power grid planning and scheduling,and is also the guarantee for the efficient economic operation of power grid.In this paper,short-term load for a province is predicted by three forecasting models.The GM(1,1)forecasting model,which is firstly established,only needs to use the historical load as a single variable.On the basis of GM(1,1),the second model GM(1,N)combines the maximum temperature,the minimum temperature,the average temperature and the relative humidity,and adds the historical load as the input.It improved the prediction precision.Based on the GM(1,N)model,a multivariable gray forecasting model GMM(1,N)modified by Markov method is proposed,and the three models are validated by the measured data of a province power grid.The results show that GMM(1,N)is the best.