Residential Load Forecasting Based On Data Mining

The residential load forecasting was analyzed based on the association rules-mining of Data-Mining and Panel Data model in this article.First,doing some work about the effect factor of urban residential and also the related characteristics.At the same time,an analysis of the residential electricity demand trends in life was considered specifically.With the development of science technology,the residents living conditions and environmental protection awareness are improving continually,and then the alternative energy of power will enter the residents living(mainly are solar,natural gas),in addition,the effect factors such as the number of urban male population,the household appliances owned(as TV,airconditioning ),capita disposable income,and the expenditure of water,gas and electricity of per capita have had a significant impact on residential electricity demand.Second,calculating the Association Rules of the related factors about residential electricity demand with R language,it can be seen that some frequent item sets about residential electricity demand was obtained through the back algorithm.Based on the results of DM,Forecasting the residential electricity demand used the Panel Data Model,and then for testing the validation and scientific of this method a comparison with some other forecasting methods has made here.