融合概率积分模型与D-InSAR的开采沉陷预计

发布时间：2018-07-07 10:52

本文选题：开采沉陷 + D-InSAR　；参考：《金属矿山》2016年02期

【摘要】：矿区地表植被多,开采沉陷速度快、量值大,所产生的地质灾害较一般性的地表沉陷严重,极易使得2景SAR影像失去相干性,造成解缠错误。针对矿区SAR影像相干性较低、下沉盆地中央相位值易丢失的情况,结合合成孔径雷达干涉差分技术(Differential interferometric synthetic aperture radar,D-In SAR)和基于遗传算法的概率积分模型,提出了一种矿区开采下沉盆地预计方法。以该方法利用矿区下沉盆地边缘一定数量的相干系数较高且下沉较明显的D-In SAR监测值和下沉盆地中央最大下沉点与拐点附近的少量观测值对某矿II3720工作面进行试验,首先利用概率积分模型反演概率积分法预计参数并采用遗传算法进行多次优化,然后利用得到的参数对该工作面下沉盆地进行模拟预计,结果表明:通过该方法得出的下沉盆地参数及下沉值与实测值较接近,有助于弥补由于矿区SAR影像干涉效果不佳而导致的预计精度不高的不足,通过少量的观测数据可较为有效地预计矿区下沉盆地,对于提高矿山开采沉陷监测与预计的精度有一定的参考价值。
[Abstract]:The mining area has more vegetation, faster mining subsidence speed and larger value, and the geological hazard is more serious than the general surface subsidence. It is easy to make two SAR images lose coherence and cause unwrapping error. In view of the low coherence of SAR images in mining areas and the easy loss of central phase values in sinking basins, the synthetic interferometric synthetic aperture radar interferometric differential SAR (SAR) and the probabilistic integral model based on genetic algorithm are combined. A prediction method of mining subsidence basin in mining area is presented. Based on the D-InSAR monitoring values with high coherence coefficient and obvious subsidence at the edge of the subsidence basin in the mining area and a few observations near the maximum subsidence point and the inflexion point in the center of the subsidence basin, this method is used to test the II3720 face in a certain mine. First, the probabilistic integral model is used to estimate the parameters and the genetic algorithm is used to optimize the parameters, and then the parameters are used to simulate and predict the subsidence basin of the working face. The results show that the parameters and subsidence values of subsidence basin obtained by this method are close to the measured values, which is helpful to make up for the deficiency of prediction accuracy caused by poor interference effect of SAR image in mining area. The subsidence basin of mining area can be predicted effectively by a small amount of observation data, which has certain reference value for improving the precision of monitoring and prediction of mining subsidence.
【作者单位】：中国矿业大学环境与测绘学院;中化地质矿山总局浙江地质勘查院;兖州煤炭股份有限公司兴隆庄煤矿;
【基金】：国家自然科学基金项目(编号:41071273) 国土环境与灾害监测国家测绘地理信息局重点实验室开放基金项目(编号:LEDM2011B07) 江苏高校优势学科建设工程基金项目(编号:SZBF2011-6-B35)
【分类号】：TD327

【相似文献】