利用Fourier级数模型预测具有非季节性周期的时间序列

Fourier Series Model in Forecasting the Non-seasonal Periodic Time Series Data

文中建立时间序列模型与Fourier级数模型的混合预测模型,使模型能够有效地对具有非季节性周期的数据进行预测。结果表明:模型能较好地反映数据的周期性,也有较准确的短期外延预测效果,但模型对奇异值的拟合和预测能力还须提高。

To establish a mixed model that contained the Time series Model and Fourier series Model.This mixed model can forecast the non-seasonal periodic data.Tthe result of its application shows： the mixed model can regress and forecast the periodic data truly,except the strange points.