基于Tukey法改进时间序列平稳性检验的分段检验法

Segmented Testing of Improved Time Series Stationarity Test Based on Tukey

在分析传统时间序列分段检验理论不足的基础上,文章提出了基于Tukey法改进时间序列平稳性检验的分段检验法。在对各段均值与协方差函数相等的检验中各构造一个t化极差统计量,减少检验犯第一类错误的概率,从而降低了平稳序列被误判为非平稳的概率,提高了分段检验方法的有效性。

Based on the analysis of deficiency of traditional time series segmented test theory,this paper proposes a segmented test method based on Tukey method to improve time series stability test. In the test of the equality of the mean and auto-covariance functions of each segment,a [t] range statistics is constructed for each to reduce the probability of making the first type of error, thus reducing the probability that the smooth sequence is misjudged as non-stationary and improving the effectiveness of the segmented test method.