灰色神经网络在电力系统短期负荷预测中的应用研究

The Application of Grey Neural Network to the Short-term Forecasting of the Load of the Electric Power System

为了有效地实现电力生产和供应,对各电网的电力负荷进行准确预测就十分必要。传统的GM(1,1)预测模型有建模数据少、计算简单和良好的短期预测能力等优点,这使得其在电力负荷的短期预测中得到了很好的应用,但是它不能有效处理电力系统的非线性问题,所以这种预测方法的预测精度不是很好。文章根据电力系统的非线性和波动性提出用灰色预测模型和神经网络理论相结合的灰色神经网络模型对电力负荷的时间序列进行短期预测。实验结果表明这种方法是可行的、有效的。

In order to make the production and the supply of power efficiently, it is very necessary to forecast the power of every power - net precisely. The traditional GM （ 1,1 ） prediction model has the advantages of less modeling data, simple calculation and good ability of short - term forecasting ere, and it has a very good application to the short - term forecasting of the power load. But it cannt dealt with the nonlinear problem of the electric power system efficiently, so the precision of forecastion of this method isnt very good . Based on the nonlinear and the fluctuate of the electric power system, this paper brings forward a grey neural network model combined with the grey theory and the neural network theory to give a short - term forecasting to the time - series of the power load. Simulation results demonstrate that the proposed method is feasible and effective.