由于地壳形变监测中观测点相对较少,用常规方法很难准确推求那些未经观测但又需要关注的点的变形量。鉴于地壳形变的特点,提出采用拟合推估方法加以解决,该方法不仅能顾及变形点现时的随机变化的影响,合理地解求已测点的变形量,还能推求未测点的形变信息。考虑到拟合推估难以确保观测噪声协方差与信号协方差之间的协调,至使拟合推估解存在偏差,为此采用自适应拟合推估的方法,即通过自适应因子调节随机信号及观测噪声之间的权比,从而提高未观测点变形量推估的准确性。最后对一实测地壳形变监测网进行形变分析,验证了上述算法的有效性。
As there are relatively few observed points in crustal deformation monitoring, it is hard to deduct the deformation of the unobserved points correctly with conventional methods. On the basis of the characteristics of the crustal deformation, the collocation method is proposed to be used, with which not only the influence of present random variation of observation points can be taken into account and the variation of the observed points can be found properly but also the deformation information of the unobserved point can be obtained. Futhermore, since the common collocation method is difficult to determine properly the weight proportion between the observation noise and stochastic signal, this paper adopts the adaptive collocation method, that is, modulating the proportion of the stochastic signal and observation noise with an adaptive factor to improve the accuracy of the deduction of the variation of the unobserved points. Finally, a measured crustal deformation monitoring network is calculated and analyzed, and the results prove the effectivity of the proposed method.