Urban Area Grid Power Load Analysis And Prediction Methods

With the development of electric power industry in our country,the deepening of electric power market reform and the increasing modernization of the power grid management,the research of power load forecasting has aroused more and more people’s attention.The accurate and efficient forecasting of power peak load forecasting can arrange the grid internal generator unit start-stop economically and reasonably,maintain the safety and stability of power grid operation,reduce the generating cost effectively,ensure the normal production and living of society,improve the economic benefit and social benefit.Urban regional power grid load analysis and forecasting methods are introduced in detail in the urban area of load classification and load characteristics,main characteristics of short term load forecasting,focusing on the linear regression,time series of the two commonly used algorithms,selected the national network of the south of the city of Tianjin power grid the typical regional network in city as an example,and developed by IBM SPSS(statistical package for the Social Sciences)platform for example case studies,the establishment of a suitable for South power grid load forecasting work of linear regression model and ARIMA model.(1)For the residents of the load forecast by four linear regression model to establish the relationship between meteorological factors(temperature,humidity,precipitation,wind speed)and residential load,test the model fitting effect,and the model’s prediction accuracy are analyzed,the model will residents class load forecasting accuracy rate is improved to more than 90%,can meet the requirements of the regional load forecasting accuracy.(2)For commercial load forecast,according to the characteristics of load changes,on reduced order processing based on the basis of the data by ARIMA model in time series algorithm analysis,to obtain a good prediction,prediction accuracy were above 90%.(3)For the industrial load forecasting.According to the load is mainly affectedby economic factors that affect the characteristics,the ARIMA model in time series methods,analysis of the relationship between the level of GDP and industrial power load,and the model fitting effect is tested,industrial negative daily load forecasting accuracy were above 90%,reaching the purpose of improving the load forecasting accuracy.