|
摘要: |
针对电力负荷预测的实际困难,提出了一种进行负荷预测的新思路,即采用节气负荷作为建模数据,并根据负荷呈现出的较为明显的时序性、周期性特点,将数据分离成趋势分量、节气周期分量,以及时间噪声及白噪声,采用双因子ARIMA模型对数据进行拟合,并以BP网络方法完成负荷预测。据此,着重论述了电力负荷预测中建模数据的选择、预处理方法及其对预测精度的影响。 |
关键词: 节气负荷 双因子ARIMA模型 BP网络 |
DOI: |
分类号:TM715 |
基金项目: |
|
Research of solar term load forecast method |
LIU Ya ZHANG Guo-zhong HE Fei
|
Abstract: |
Because of the difficulty in power load forecast,a new c oncept of lo ad forecast is put forward,which takes solar term loads as model ing data.According to obvious chronological feature and periodicity of power lo ad,the data is separated into trend factor,solar term period factor,calen-da r jam and random jam.The dual factorial ARIMA model is used to fit the data and the BP network to forecast electric load.Based on it,the selection of modelin g data,data pretreatment and their influences on forecast precision are discus sed. |
Key words: solar term load,dual factorial ARIMA model,BP network |