基于改进的灰色理论在中长期电力负荷预测中的应用

Prediction model for medium and long term electric load based on improved grey theory

面对中长期电力负荷预测"小样本"、"少信息"、"非线性"等特点[1],灰色预测模型在电力负荷预测中起决定性的作用。该文论述了基本GM(1,1)模型、一次平滑法的GM(1,1)模型。针对以上两种模型的缺点和不足,通过对初始数据的二次平滑处理,又提出了改进的灰色预测模型——二次平滑法的GM(1,1)模型。通过算例检验与典型的实例研究上述3种灰色模型在中长期负荷预测中的应用。结果表明,和前面两种预测模型相比较,二次平滑法的GM(1,1)模型在电力系统电量的实际预测中更精确,误差更小。

The face of medium and long term power load forecasting＂small sample＂,＂little informa-tion＂,＂non-linear＂etc[1] , gray prediction model play a decisive role in the power load forecasting. This paper discusses a basic GM （1,1） model, a smoothing method of GM （1,1） model.For the short-comings and deficiencies of these two models, through the secondary smoothing the initial data, and pro-posed an improved gray prediction model -quadratic smoothing method GM（1,1） model.Studies of the three gray model in medium and long term load forecasting inspection by an example of typical examples. The results show that prediction model compared with the previous two, quadratic smoothing method GM（1,1） model is more accurate in predicting the actual electricity grid, the error is smaller.