基于EMD理论的短期负荷预测

Short-term load forecasting based on EMD theory

提出将经验模态分解法(empirical mode decomposition,EMD)与自回归动平均(autoregressive and moving average,ARMA)时序分析相结合的短期负荷预测方法,利用EMD算法对负荷序列进行分解,再对分解得到的每一个基本模式分量(intrinsic mode function,IMF)分别进行ARMA预测,以改善短期负荷预测的效果。以我国某地区2008年夏季负荷高峰段的用电负荷预测为例,应用提出的预测方法对各负荷分量进行单独预测,得到的仿真结果表明,采用该方法进行短期负荷预测将显著改善预测的精度。

This paper presents the short-term load forecasting method which combines the EMD theory with the ARMA. This method decomposes the load sequence into intrinsic mode function with EMD algorithm,predicts the separate IMF with the ARMA,which can improve the effectiveness of the short-term load forecasting. This paper takes the load forecasting of 2008 summer load peak section of some area as an example and separately predicts the load components applying the forecasting method which proposed by the predictive model. The simulation results obtained show that the forecast accuracy will be significantly improved by using the model.