大坝监测分析中的时间序列叠合模型

叙述了用时间序列叠合模型建立大坝监测预报模型的2种基本方法：统计-时间序列模型和确定性时序叠合模型。通过统计软件SPSS实现2种时序叠合模型的实例表明，在环境量记录完整的情况下，统计-时间序列模型比单纯的统计回归模型精度要高；在缺少环境量监测或与环境量的因果关系不明显的情况下，用确定性叠合模型分析和预报大坝监测效应量的测值，精度也很高，并能够分离出趋势性分量、周期分量和随机平稳分量，分析的结果简单明了，是大坝监测分析中一种比较好的方法。

The Time-series Superposition Model for Dam Monitoring Analysis

This paper describes two kinds of basic methods to construct a dam monitoring forecast model with the time-series superposition model:statistic time-series model and deterministic superposition model.An example of the two kinds of time- series superposition models realized with the SPSS statistic software shows that:under the condition of complete environmental observation data,the accuracy of the statistic time-series model is higher than that of the single statistic regression model,and the accuracy of the deterministic superposition model is also high in defect of environmental observation data or without clear consequence of them;at the same time,the trend component,periodic component and random factor component can be separated.The analytic result is simple and clear.It is proved fit for dam monitoring analysis.