改进的残差修正DGM(1,1)模型在煤矿地表沉降预测中的应用

Application of improved residual correction DGM (1,1) model in the prediction of coal mine surface subsidence

针对传统大地测量技术难以实施大范围矿区形变监测以及传统开采沉陷预测方法在实际应用中存在的不足,提出时序合成孔径雷达干涉测量(InSAR)和改进的离散灰色马尔科夫模型相结合的方法对矿区地表形变进行监测和预测。使用鄂尔多斯市某矿区哨兵1号(Sentinel-1)数据提取地表形变监测结果,并与同时期的全球定位系统(GPS)监测数据进行对比。将沉陷监测结果作为原始序列建立残差修正的离散灰色马尔科夫模型,预测矿区沉降。研究结果表明:研究区域地表形变明显,残差修正的离散灰色马尔科夫模型能较为准确地预测矿区沉降结果。

In view of the difficulties of traditional geodetic technology in monitoring the deformation of large mining areas and the shortcomings of traditional mining subsidence prediction methods in practical application, a method combining time series InSAR and improved discrete grey Markov model was proposed to monitor and predict the surface deformation of mining areas. The Sentinel-1 data of a mining area in Ordos City was applied to extract the surface deformation monitoring results and compare them with the GPS monitoring data of the same period. The subsidence monitoring results were taken as the original sequence to establish the discrete grey Markov model with residual correction to predict the subsidence of the mining area. The results showed that the surface deformation in the study area was obvious, and the discrete grey Markov model with residual correction could accurately predict the subsidence of the mining area.