指数平滑法及其在负荷预测中的应用

Exponential Smoothing Method and Its Application to Load Forecasting

指数平滑法是电力系统负荷预测的主要方法之一,该方法的准确性取决于平滑系数α.对采用厚近薄远原则与远近相同原则优选α进行对比研究,结果表明采用厚近薄远原则优选α有更好的结果.在此基础上,结合相关分析,给出了厚近薄远的具体方案,并给出了负荷预测的具体实例.

Exponential smoothing method is one of the main load forecasting methods for power system;the accuracy of the method depends on smoothing coefficient α.In this paper,a study of how to seek the best α is given.The results show that the use of principle of valueing near errors and contempting far errors can get a better model.On this basis,combined with correlative analysis,a method of how to value near errors and contempt far errors is proposed.A practical example of load forecasting is also shown here.