针对永久散射体差分干涉测量(PSIn SAR)算法流程,发展了基于周期函数模型的空间维解缠方法,并将其应用于矿区时间序列地表变形反演。通过在研究区域内安装人工角反射器(CR),将CR点上计算所得的周期模型参数分量作为整个网络的约束,通过空间约束平差以实现空间维解缠。选取了河南省境内白沙水库附近的煤矿密集区为主要研究区域,采用周期函数模型对矿区线性及非线性形变分量进行模拟,反演了2007年2月—2010年2月的时间序列形变场,并采用研究区域内的水准实测数据作为外部验证数据。实验结果表明:白沙水库周围区域存在着较为明显的沉降,在煤矿分布区域内累积最大沉降量超过了10 cm。沉降区域内以线性沉降趋势为主,非线性沉降较为缓慢,仅在水库的西南方向较为明显。应用已有的水准点实测形变值对实验结果进行验证分析,结果表明该方法精度可达约±2.1 mm,证实了本文采用的方法在矿区地表时序形变反演中的可行性和可靠性,对预防过度采矿导致的矿区塌陷具重要现实意义。 Aiming at the PSIn SAR algorithm flow, a space dimension unwrapping method based on periodic function model was developed and applied to time series surface deformation inversion of mining area. By installing an artificial corner reflector (CR) in the study area, the parameters of the periodic model calculated from the CR point are used as the constraints of the entire network, and the space constraint solution is used to achieve the spatial dimension unwrapping. Taking the coal concentration area near Baisha reservoir in Henan Province as the main research area, the linear and nonlinear deformation components of the mining area are simulated by the periodic function model, and the time series deformation field is retrieved from February 2007 to February 2010. The survey data from the survey area are used as external verification data. The experimental results show that there is obvious settlement in the surrounding area of ​​Baisha reservoir, and the maximum cumulative settlement in the coal mine distribution area exceeds 10 cm. The tendency of linear settlement in the subsidence area is dominated, and the nonlinear subsidence is relatively slow, which is only obvious in the southwest of the reservoir. The experimental results are validated by using the measured deformation of leveling point. The results show that the accuracy of this method can reach about ± 2.1 mm, which proves the feasibility and reliability of the method adopted in this study in the inversion of surface sequence deformation in mining area. It is of great practical significance to prevent mining subsidence caused by over-mining.