为了提高有限样本或小样本情况下时间序列均值和方差函数的确定精度,以保证时序建模、分析和预测精度,提出一种确定序列趋势项的移动多点平均方法.该方法能够得到时间序列均值中非周期部分,结合样本周期图法得到的周期项,可综合得到其均值函数,并可进一步得到时序的标准差函数.MonteCarlo模拟对比结果表明:该方法不仅能够保证时间序列段内分析精度,而且能够有效提高时间序列的预报精度. In order to improve the accuracy of time series averaging and variance function in the case of finite sample or small sample, in order to ensure the accuracy of timing modeling, analysis and prediction, a moving multi-point averaging method to determine the trend of sequence is proposed. In the aperiodic part of the mean and the periodic term obtained by the sample periodogram method, the mean function can be obtained synthetically, and the standard deviation function of the sequence can be further obtained. The simulation results of Monte Carlo simulation show that this method can not only ensure the analysis accuracy , But also can effectively improve the forecast accuracy of time series.