基于EMD-分形理论的短期电力负荷预测

Short-term power load forecasting based on fractal Empirical Mode Decomposition theory

电力负荷预测是电力系统规划的重要组成部分。为了使电力系统安全经济平稳的运行,由此特别需要精确的电力负荷预测方法。为了实现更好负荷预测方法,文中将经验模态分解(EMD)与新兴的电力负荷预测模型分形理论相结合,提出了EMD-分形负荷预测模型。为了证明此方法的有效性,文中将这种新的预测模型跟分形预测模型和BP神经网络预测模型相比较。最终通过仿真算例说明了本文提出的这种新型预测方法精度更高,几乎所有的误差都在2%以下,预测结果更好,可以很好的应用在电力系统负荷预测中。

Power load forecasting is an important part of power system. To make the power system stability, therefore special needs accurate load forecasting method. In this paper, in order to achieve a better load forecasting mode, combine the Empirical Mode Decomposition（EMD） and fractal theory and improve the EMD fractal load forecasting model. In order to prove the effectiveness of this method, the paper use this new forecasting model compared with fractal forecasting model and BP neural network forecasting model. Finally the simulate example illustrates this new forecasting method proposed higher precision and better forecasting. Almost all of the errors are below 2%. The result proved it can be a good application in electric power system load forecasting.