将地球物理模型估计的形变量与GPS几何观测量结合进行形变分析时,由于受各种因素的影响,地球物理模型参数难免存在不可靠现象,直接应用这种不可靠信息进行参数估计,会影响参数解的可靠性。将地球物理模型估计的形变信息作为先验信息,与几何观测量相结合进行形变分析,并提出一种能够有效修正不可靠先验信息的基于稳健轮回搜索的Bayes估计方法。考虑到异常形变点或大的观测误差对轮回搜索结果存在较大影响,又提出用高崩溃污染率的稳健估计法来确定能够代表块体均匀应变的一组相对稳定的点,由这些相对稳定点组计算搜索函数,确保先验信息的可靠性。最后利用一实测GPS监测网进行计算,结果表明稳健轮回搜索可以有效地对失真先验信息进行修正,提高模型参数解的精度。
When combining geophysical information with geodetic observation for crustal deformation analysis there may be unreliable parameters.The Bayes estimation based on robust cyclic searching is proposed,in which the deformation predicted by geophysical model as priori information in geodetic(GPS) measurements processing is used and the cyclic searching is adopted to modify improper priori information.Taking into account the abnormal observation station moving or outliers in measurements may influences the searching results,we calculate searching functions from selected stable stations by robust cyclic searching.The new method is applied in a GPS monitoring net.It is shown that the robust cyclic searching is effective in modifying priori information and improving the accuracy of model parameter.