论文部分内容阅读
笔者应用逐步回归和契氏周期分析法,对宽城台目视水管倾斜仪连续五年观测资料进行处理:较好地排除了各种因素对观测资料的影响。回归计算结果,复相关系数γ_(EW)=0.99,γ_(NS)=0.93,中误差M_(EW)=±5.3、M_(NS)=±4.6微米,契氏法拟合结果,前三阶分量占主导优势,反映了该台资料的卓越周期余差平稳,五年平均中误差为M_(EW)=5.6,M_(NS)=3.5微米,与回归计算精度相当。本文还研究了叫归计算中采用的诸物理因素和卓越周期之间关系以及干扰机制。最后还讨论了两种分析法的特点和应用。
The application of stepwise regression and Cheney ’s period analysis method, the wide city of Taiwan as the water pipe tilt meter for five consecutive years of observation data processing: a better rule out the impact of various factors on the observed data. The results of regression calculation showed that the complex correlation coefficient γ_ (EW) = 0.99, γ_ (NS) = 0.93, the error M_ (EW) = ± 5.3, M_ (NS) The component predominance reflects the excellent period residual error of the station data. The average error of the five years is M_ (EW) = 5.6 and M_ (NS) = 3.5 μm, which is equivalent to the precision of regression calculation. The paper also studies the relationship between physical factors used in computing and epochs and the mechanism of interference. Finally, it also discusses the characteristics and applications of the two analytic methods.