时间序列分析论文

St2StSt指数平滑法无论是一次指数平滑或者二次、三次指数平滑值都不适合直接作为预测值,但是都能修匀时间序列,于是可以建立如下的三次指数平滑趋势预测模型。首先:选取α=0.7,α=0.75,α=0.8,α=0.85,α=0.90,α=0.95对原始数据进行指数平滑[13]。经过EXCLE2003指数平滑分析,当阻尼系数为0.05,即平滑系数α为0.95时最适合做指数平滑预测:如下表3表3三次指数平滑结果自一次平滑二次平滑三次平滑然绵阳市旅游总收年份序入状况(亿元)α=0.95号199718.18.18.18.1199829.39.249.1839.128851999310.610.53210.4645510.397772000422.6822.072621.492220.937482001521.4821.5096321.5087621.480192002625.5625.3574825.1650524.98082003728.328.1528728.0034827.852352004836.0335.6361435.2545134.88442005945.0744.5983144.1311243.6687820061063.562.5549261.6337360.7354820071180.979.9827579.0652978.148820081240.242.1891444.0329545.7387420091347.847.5194647.3451347.2648120101465.264.3159763.4674362.657320111597.295.555893.9513892.38668根据上述公式选取t=15、α=0.95建立指数平滑三次曲线趋势预测模型为:∧=97.20+0.21l+0.36l2其中=97.20=0.21=0.36Yt+latbtct8