Application of self-organizing combination forecasting method in power load forecast
Wei Sun, Xing Zhang · 2007
According to the load properties of electric power, four kinds of component forecasting models are chosen and a new combination forecasting model based on Self-organizing data mining algorithm is introducted in this paper. The forecasted results of each component forcasting models are used as the input of self-organizing data mining algorithm, and the output are the results of combination forecasting. In order to vertify the validity and maneuverability of the model, a load forecasting example is given and the result show that this model can improve the forecasting ability remarkably when comparing to optimal combination forecasting and artificial neural network combination forecasting.