Household Electric Power Load Forecasting based on A Novel Combining Model
The individual household electricity consumption is major part of the city in the electricity market.The accurate prediction of household power load is very important for power sector to reasonable decision and reduced operating costs of Electricity Market.In this paper, a novel combining model is proposed which capable of utilizing advantages of SARIMA and SVM model respectively, and the weight variables of α1 and α2 are proposed which optimized by particle swarm optimization (PSO) algorithm for showing the relationship of linear and nonlinear part in real power load dataset.The electric dataset of household power consumption is used to examine performance of the developed combined method.Experimental results indicate that the novel model can obtains better robust and precise than conventional models of SARIMA and SVM separately.
support vector machines seasonal autoregressive integrated moving average combined model time series
TAO Ma Yukai Yao Fen WVang Xiaoyun Chen
School of Information Science and Engineering,Lanzhou University,Lanzhou 730000,China Computer and Communication,Lanzhou University of Technology,Lanzhou 730050,China School of Mathematical and Computer Science,Ningxia Teachers University,Guyuan,Ningxia 756000 China
国内会议
金华
英文
1-11
2015-10-30（万方平台首次上网日期，不代表论文的发表时间）