融合D-InSAR技术和SVM算法的城口锰矿开采沉陷预计模型
|
摘要
由于传统矿山沉陷监测技术和沉陷预计方法难以进行有机结合,导致矿山开采沉陷预计精度较低。以重庆城口锰矿为例,将D-In SAR技术与SVM算法进行有机结合,采用由D-In SAR(Differential synthetic aperture radar)技术获取的矿区开采沉陷监测值作为支持向量机(Support vector machine,SVM)算法的训练样本,得到了SVM算法参数,构建了开采沉陷预计的SVM模型。研究表明:预计值与D-In SAR实测值的最大绝对误差为12 mm,最小绝对误差为0.06 mm,平均绝对误差为7.08 mm,最大相对误差为4.2%,最小相对误差为0.14%,平均相对误差为2.3%,表明所构建的SVM开采沉陷预计模型精度较高。
Due to the combination of traditional mining subsidence monitoring techniques and mining subsidence mining prediction algorithms is difficult,which caused the precise of mining subsidence prediction is low. In order to solve the problem,taking Chengkou Manganese Deposit in Chongqing City as the study example,the D-In SAR(differential synthetic aperture radar) technique and SMV(support vector machine) algorithm are integrated,the mining subsidence monitoring results obtained by D-In SAR techniques is taken as the training sample data of SVM algorithm,the parameters of SVM algorithm are obtained,and the SVM mining subsidence prediction model of the mining area is established. The study results show that: the maximum absolute error between the actual monitoring data obtained by D-In SAR technique and SVM prediction is 12 mm,the minimum absolute error between the actual monitoring data obtained by D-In SAR technique and SVM prediction model is 0. 06 mm,the average absolute error between the actual monitoring data obtained by D-In SAR technique and SVM prediction model is 7. 08 mm,the maximum relative error of them is 4. 2%,the minimum relative error of them is 0. 14%,the average relative error of them is 2. 3%,which show that the precise of the SVM prediction model established by the D-In SAR actual monitoring data is consistent with the actual monitoring data.
Due to the combination of traditional mining subsidence monitoring techniques and mining subsidence mining prediction algorithms is difficult,which caused the precise of mining subsidence prediction is low. In order to solve the problem,taking Chengkou Manganese Deposit in Chongqing City as the study example,the D-In SAR(differential synthetic aperture radar) technique and SMV(support vector machine) algorithm are integrated,the mining subsidence monitoring results obtained by D-In SAR techniques is taken as the training sample data of SVM algorithm,the parameters of SVM algorithm are obtained,and the SVM mining subsidence prediction model of the mining area is established. The study results show that: the maximum absolute error between the actual monitoring data obtained by D-In SAR technique and SVM prediction is 12 mm,the minimum absolute error between the actual monitoring data obtained by D-In SAR technique and SVM prediction model is 0. 06 mm,the average absolute error between the actual monitoring data obtained by D-In SAR technique and SVM prediction model is 7. 08 mm,the maximum relative error of them is 4. 2%,the minimum relative error of them is 0. 14%,the average relative error of them is 2. 3%,which show that the precise of the SVM prediction model established by the D-In SAR actual monitoring data is consistent with the actual monitoring data.
引文
[1]王磊,张鲜妮.基于单视线向D-In SAR技术提取开采沉陷移动与变形方法研究[J].岩石力学与工程学报,2017,36(4):865-873.Wang Lei,Zhang Xianni.Retrieving surface movement and deformation induced by coal mining based on a single D-In SAR pair[J].Chinese Journal of Rock Mechanics and Engineering,2017,36(4):865-873.
[2]Singleton A,Li Z,Hoey T,et al.Evaluating sub-pixel offset techniques as an alternative to D-In SAR for monitoring episodic landslide move ments in vegetated terrain[J].Remote Sensing of Environment,2014,147(9):133-144.
[3]Schl9gel R,Doubre C,Malet J P,et al.Landslide deformation monitoring with ALOS/PALSAR imagery:a D-In SAR geomorphological interpretation method[J].Geomorphology,2015(2):314-330.
[4]肖先煊,许强,刘家春,等.IBIS-L在滑坡地表变形监测中的应用——以宁南县白水河滑坡为例[J].长江科学院院报,2015,32(8):45-50.Xiao Xianxuan,Xu Qiang,Liu Jiachun,et al.Application of IBIS-L to the monitoring of landslide deformation:case study of Baishuihe landslide in Southwest China[J].Journal of Yangtze River Scientific Research Institute,2015,32(8):45-50.
[5]孟万利,蔡来良,杨望山,等.矿区开采沉陷预计的Geomagic法[J].金属矿山,2017(1):126-131.Meng Wanli,Cai Liangliang,Yang Wangshan,et al.Mining subsidence prediction method based on Geomagic[J].Metal Mine,2017(1):126-131.
[6]陈涛,郭广礼,朱晓峻,等.利用果蝇算法反演概率积分法开采沉陷预计参数[J].金属矿山,2016(6):185-188.Chen Tao,Guo Guangli,Zhu Xiaojun,et al.Mining subsidence parameters inversion of the probability integral method based on fruit flies algorithm[J].Metal Mine,2016(6):185-188.
[7]刘哲.基于D-In SAR的煤矿开采沉陷变形监测技术研究与应用[D].淮南:安徽理工大学,2016.Liu Zhe.Research and Application of Coal Mining Subsidence Deformation Monitoring Technology Based on D-In SAR[D].Huainan:Anhui University of Science and Technology,2016.
[8]田辉.D-In SAR技术在盘锦地区地面沉降监测中的应用研究[D].长春:吉林大学,2014.Tian Hui.Application of D-In SAR Technology in Land Subsidence Monitoring in Panjin Area[D].Changchun:Jilin University,2014.
[9]赵飞.基于支持向量机的煤矿开采沉陷预测模型研究[J].矿山测量,2015(6):27-29.Zhao Fei.Study on prediction model of coal mining subsidence based on support vector machine[J].Mining Survey,2015(6):27-29.
[10]拓万兵,姜伟,吴凤民.基于支持向量机的开采沉陷预计参数选取研究[J].中国矿业,2015(2):114-116.Tuo Wanbing,Jiang Wei,Wu Fengmin.Study on the selection of prediction parameters on mining subsidence based on support vector machine[J].China Mining Magazine,2015(2):114-116.
[11]李怀展,查剑锋,米丽倩.基于卡尔曼滤波的D-In SAR和水准监测数据融合方法研究[J].大地测量与地球动力学,2015,35(3):472-476.Li Huaizhan,Zha Jianfeng,Mi Liqian.Study on D-In SAR and leveled data fusion method based on Kalman filter[J].Journal of Geodesy and Geodynamics,2015,35(3):472-476.
[12]杨俊凯,邓喀中,赵伟颖,等.基于D-In SAR和概率积分法的矿区沉降预计[J].煤炭技术,2015(4):108-110.Yang Junkai,Deng Kazhong,Zhao Weiying,et al.Development prediction of mining area based on D-In SAR and probability integral method[J].Coal Science and Technology,2015(4):108-110.
[13]徐超,徐贵娃,马可.基于支持向量机的静压管桩单桩极限承载力预测[J].露天采矿技术,2015(8):83-85.Xu Chao,Xu Guiwa,Ma Ke.Fast support capacity of single pile of static pressure pile based on support vector machine[J].Opencast Mining Technology,2015(8):83-85.
[14]张鲜妮,王磊.基于支持向量机的建筑物釆动损害预测模型[J].矿业安全与环保,2016,43(6):40-44.Zhang Xianni,Wang Lei.Building dynamic damage prediction model based on support vector machine[J].Mining Safety and Environmental Protection,2016,43(6):40-44.
[15]田馨,廖明生.In SAR技术在监测形变中的干涉条件分析[J].地球物理学报,2013,56(3):812-823.Tian Xin,Liao Mingsheng.Analysis of interference conditions of InSAR technology in monitoring deformation[J].Chinese Journal of Geophysics,2013,56(3):812-823.
[16]敖萌,张勤,赵超英,等.改进的CR-In SAR技术用于四川甲居滑坡形变监测[J].武汉大学学报:信息科学版,2017,42(3):377-383.Ao Meng,Zhang Qin,Zhao Chaoying,et al.An improved CR-In SAR technology used for deformation monitoring in Jiaju landslide,Sichuan[J].Geomatics and Information Science of Wuhan University,2017,42(3):377-383.
[17]康日斐,吴泉源,王菲,等.基于D-In SAR技术的龙口矿区地表沉降遥感监测研究[J].土壤通报,2016,47(5):1049-1055.Kang Rifei,Wu Quanyuan,Wang Fei,et al.Remote sensing monitoring of surface subsidence in Longkou mining area based on D-InSAR[J].Chinese Journal of Soil Science,2016,47(5):1049-1055.
[18]李焱磊,梁兴东,丁赤飚.机载差分干涉SAR双轨法和三轨法的误差比较分析[J].雷达学报,2013(3):326-333.Li Yanlei,Liang Xingdong,Ding Chibiao.Comparative analysis of error of dual-track method and three-track method for airborne differential interferometric SAR[J].Acta Radar Sinica,2013(3):326-333.
[19]王素英,张喆,赵悦品.果蝇算法融合SVM的开采沉陷预测模型[J].煤矿安全,2014,45(5):43-46.Wang Suying,Zhang Zhe,Zhao Yuepin.Prediction model of mining subsidence for frog flyer fusion SVM[J].Safety in Coal Mines,2014,45(5):43-46.
[20]温廷新,孙红娟,徐波,等.矿区采空塌陷危险性预测的RSSVM模型[J].中国安全科学学报,2015,25(10):16-21.Ying Tingxin,Sun Hongjuan,Xu Bo,et al.RS-SVM model for prediction of subsidence risk in mining area[J].Journal of Chinese Society for Safety Science,2015,25(10):16-21.
[21]黄禄文,刘海卿.建筑物基础沉降径向基神经网络预测[J].辽宁工程技术大学学报:自然科学版,2016,35(8):836-840.Huang Luwen,Liu Haiqing.Foundation settlement of building and RBF neural networks prediction[J].Journal of Liaoning Technical University:Natural Science Edition,2016,35(8):836-840.
[2]Singleton A,Li Z,Hoey T,et al.Evaluating sub-pixel offset techniques as an alternative to D-In SAR for monitoring episodic landslide move ments in vegetated terrain[J].Remote Sensing of Environment,2014,147(9):133-144.
[3]Schl9gel R,Doubre C,Malet J P,et al.Landslide deformation monitoring with ALOS/PALSAR imagery:a D-In SAR geomorphological interpretation method[J].Geomorphology,2015(2):314-330.
[4]肖先煊,许强,刘家春,等.IBIS-L在滑坡地表变形监测中的应用——以宁南县白水河滑坡为例[J].长江科学院院报,2015,32(8):45-50.Xiao Xianxuan,Xu Qiang,Liu Jiachun,et al.Application of IBIS-L to the monitoring of landslide deformation:case study of Baishuihe landslide in Southwest China[J].Journal of Yangtze River Scientific Research Institute,2015,32(8):45-50.
[5]孟万利,蔡来良,杨望山,等.矿区开采沉陷预计的Geomagic法[J].金属矿山,2017(1):126-131.Meng Wanli,Cai Liangliang,Yang Wangshan,et al.Mining subsidence prediction method based on Geomagic[J].Metal Mine,2017(1):126-131.
[6]陈涛,郭广礼,朱晓峻,等.利用果蝇算法反演概率积分法开采沉陷预计参数[J].金属矿山,2016(6):185-188.Chen Tao,Guo Guangli,Zhu Xiaojun,et al.Mining subsidence parameters inversion of the probability integral method based on fruit flies algorithm[J].Metal Mine,2016(6):185-188.
[7]刘哲.基于D-In SAR的煤矿开采沉陷变形监测技术研究与应用[D].淮南:安徽理工大学,2016.Liu Zhe.Research and Application of Coal Mining Subsidence Deformation Monitoring Technology Based on D-In SAR[D].Huainan:Anhui University of Science and Technology,2016.
[8]田辉.D-In SAR技术在盘锦地区地面沉降监测中的应用研究[D].长春:吉林大学,2014.Tian Hui.Application of D-In SAR Technology in Land Subsidence Monitoring in Panjin Area[D].Changchun:Jilin University,2014.
[9]赵飞.基于支持向量机的煤矿开采沉陷预测模型研究[J].矿山测量,2015(6):27-29.Zhao Fei.Study on prediction model of coal mining subsidence based on support vector machine[J].Mining Survey,2015(6):27-29.
[10]拓万兵,姜伟,吴凤民.基于支持向量机的开采沉陷预计参数选取研究[J].中国矿业,2015(2):114-116.Tuo Wanbing,Jiang Wei,Wu Fengmin.Study on the selection of prediction parameters on mining subsidence based on support vector machine[J].China Mining Magazine,2015(2):114-116.
[11]李怀展,查剑锋,米丽倩.基于卡尔曼滤波的D-In SAR和水准监测数据融合方法研究[J].大地测量与地球动力学,2015,35(3):472-476.Li Huaizhan,Zha Jianfeng,Mi Liqian.Study on D-In SAR and leveled data fusion method based on Kalman filter[J].Journal of Geodesy and Geodynamics,2015,35(3):472-476.
[12]杨俊凯,邓喀中,赵伟颖,等.基于D-In SAR和概率积分法的矿区沉降预计[J].煤炭技术,2015(4):108-110.Yang Junkai,Deng Kazhong,Zhao Weiying,et al.Development prediction of mining area based on D-In SAR and probability integral method[J].Coal Science and Technology,2015(4):108-110.
[13]徐超,徐贵娃,马可.基于支持向量机的静压管桩单桩极限承载力预测[J].露天采矿技术,2015(8):83-85.Xu Chao,Xu Guiwa,Ma Ke.Fast support capacity of single pile of static pressure pile based on support vector machine[J].Opencast Mining Technology,2015(8):83-85.
[14]张鲜妮,王磊.基于支持向量机的建筑物釆动损害预测模型[J].矿业安全与环保,2016,43(6):40-44.Zhang Xianni,Wang Lei.Building dynamic damage prediction model based on support vector machine[J].Mining Safety and Environmental Protection,2016,43(6):40-44.
[15]田馨,廖明生.In SAR技术在监测形变中的干涉条件分析[J].地球物理学报,2013,56(3):812-823.Tian Xin,Liao Mingsheng.Analysis of interference conditions of InSAR technology in monitoring deformation[J].Chinese Journal of Geophysics,2013,56(3):812-823.
[16]敖萌,张勤,赵超英,等.改进的CR-In SAR技术用于四川甲居滑坡形变监测[J].武汉大学学报:信息科学版,2017,42(3):377-383.Ao Meng,Zhang Qin,Zhao Chaoying,et al.An improved CR-In SAR technology used for deformation monitoring in Jiaju landslide,Sichuan[J].Geomatics and Information Science of Wuhan University,2017,42(3):377-383.
[17]康日斐,吴泉源,王菲,等.基于D-In SAR技术的龙口矿区地表沉降遥感监测研究[J].土壤通报,2016,47(5):1049-1055.Kang Rifei,Wu Quanyuan,Wang Fei,et al.Remote sensing monitoring of surface subsidence in Longkou mining area based on D-InSAR[J].Chinese Journal of Soil Science,2016,47(5):1049-1055.
[18]李焱磊,梁兴东,丁赤飚.机载差分干涉SAR双轨法和三轨法的误差比较分析[J].雷达学报,2013(3):326-333.Li Yanlei,Liang Xingdong,Ding Chibiao.Comparative analysis of error of dual-track method and three-track method for airborne differential interferometric SAR[J].Acta Radar Sinica,2013(3):326-333.
[19]王素英,张喆,赵悦品.果蝇算法融合SVM的开采沉陷预测模型[J].煤矿安全,2014,45(5):43-46.Wang Suying,Zhang Zhe,Zhao Yuepin.Prediction model of mining subsidence for frog flyer fusion SVM[J].Safety in Coal Mines,2014,45(5):43-46.
[20]温廷新,孙红娟,徐波,等.矿区采空塌陷危险性预测的RSSVM模型[J].中国安全科学学报,2015,25(10):16-21.Ying Tingxin,Sun Hongjuan,Xu Bo,et al.RS-SVM model for prediction of subsidence risk in mining area[J].Journal of Chinese Society for Safety Science,2015,25(10):16-21.
[21]黄禄文,刘海卿.建筑物基础沉降径向基神经网络预测[J].辽宁工程技术大学学报:自然科学版,2016,35(8):836-840.Huang Luwen,Liu Haiqing.Foundation settlement of building and RBF neural networks prediction[J].Journal of Liaoning Technical University:Natural Science Edition,2016,35(8):836-840.