基于Elman神经网络的负荷预测研究

Research on Load Forecasting Based on Elman Neural Network

电力系统负荷预测是电力生产部门的重要工作之一,其负荷变化具有明显的周期性,文章采用Elman神经网络与BP神经网络建立模型,提出了一种基于神经网络的负荷预测方法。对某电网实际历史数据进行仿真预测,经研究发现,Elman模型具有收敛速度快、预测精度高的特点,同时表明利用Elman回归神经网络建模对某电网负荷进行预测是完全可行的,在负荷预测领域有着较好的应用前景。

Power system load forecasting is one of the main tasks of power companies. The load changes in a periodical manner. In this paper, we will try to use an ANN （artificial neural network） approach to forecast load changes. We will put this approach to test with some historical data from a power net work. The result is quite satisfying, Elman model is efficient and accurate. Therefore we could use the Elman neural network to forecast load changes of power net work, put it into practical use.