基于回归分析及卡尔曼滤波的闽赣断裂带垂直形变预测
|
摘要
断裂带的活动性与地震活动有着密切的关系,断裂带垂直形变监测对捕捉地震前兆异常情况、预测地震危险性均有着非常重要的作用。基于闽赣断裂带5年跨断层水准观测数据,首先,通过拟合残差、F检验置信度水平综合判断各监测场地的多项式拟合阶数,构建了拟合效果较优的多项式回归模型;其次,引入卡尔曼滤波理论,分析并确定滤波先验信息,建立了垂直形变的卡尔曼滤波模型;最后,基于建立的两种模型对闽赣断裂带垂直形变进行预测分析。结果表明:当测段高差前后两期出现较大波动时,多项式回归模型预测精度略高于卡尔曼滤波模型;反之,监测场地测段高差变化较为平缓时,采用卡尔曼滤波模型预测精度优于1mm,在实际应用中,可以综合应用两种模型进行垂直形变预测,以提高预测效果。
The activity of fault zone is closely related to earthquake activity,and also the prediction of vertical deformation of fault zone plays an important role in predicting earthquake precursor anomalies and earthquake hazards. According to the five years of cross-fault horizontal observation data of Fujian-Jiangxi fault zone,in this paper,firstly,the polynomial fitting order of each monitoring site was comprehensively determined by fitting residual error and F test confidence level,and the polynomial regression model with the best fitting effect was constructed.Secondly,the Kalman filtering theory was introduced,with analyzing and determining the prior information of filtering,we established Kalman filtering model of vertical deformation. Finally,the vertical deformation of the FujianJiangxi fault zone was predicted and analyzed based on the above two models. The results showed that the prediction accuracy with polynomial regression model is slightly higher than that with Kalman filtering model when there is a large fluctuation in the observation value. On the contrary,when the height difference is relatively mild,the prediction accuracy with Kalman filtering model is better than 1 mm. In practical application,two models can be used to predict vertical deformation comprehensively,so as to improve the prediction effect.
The activity of fault zone is closely related to earthquake activity,and also the prediction of vertical deformation of fault zone plays an important role in predicting earthquake precursor anomalies and earthquake hazards. According to the five years of cross-fault horizontal observation data of Fujian-Jiangxi fault zone,in this paper,firstly,the polynomial fitting order of each monitoring site was comprehensively determined by fitting residual error and F test confidence level,and the polynomial regression model with the best fitting effect was constructed.Secondly,the Kalman filtering theory was introduced,with analyzing and determining the prior information of filtering,we established Kalman filtering model of vertical deformation. Finally,the vertical deformation of the FujianJiangxi fault zone was predicted and analyzed based on the above two models. The results showed that the prediction accuracy with polynomial regression model is slightly higher than that with Kalman filtering model when there is a large fluctuation in the observation value. On the contrary,when the height difference is relatively mild,the prediction accuracy with Kalman filtering model is better than 1 mm. In practical application,two models can be used to predict vertical deformation comprehensively,so as to improve the prediction effect.
引文
[1]张国民,傅征祥,王晓青,等.确定全国地震重点监视防御区的研究[J].中国地震,2006,22(3):209-221.ZHANG Guomin,FU Zhengxiang,WANG Xiaoqing,et al. Study on determination of the national significant seismic monitoring and protection regions[J]. Earthquake Research in China,2006,22(3):209-221.(in Chinese)
[2] Cheng Wanzheng,Yang Yongling. Deformation rate changes of tectonic belts along boundaries of Yunnan Sichuan block and relation to grouped strong earthquakes[J]. Journal of Geodesy and Geodynamics,2003,23(b12):101-107.
[3]黄卿团,郑韶鹏.福建东南沿海及邻区活动断裂的微地貌研究[J].地球物理学进展,2006,21(4):1099-1107.HUANG Qingtuan,ZHENG Shaopeng. Micro-relief research on active fault in the coast of southeast Fujian and adjacent area[J]. Progress in Geophysics,2006,21(4):1099-1107.(in Chinese)
[4]苏琴,杨永林,向和平.汶川8. 0级地震前后鲜水河断裂带断层活动特征分析[J].四川地震,2012(142):24-29.SU Qin,YANG Yonglin,XIANG Heping. Seismicity character analysis on the Xianshuihe faults before and after the 2008 Wenchuan MS8. 0Earthquake[J]. Earthquake Research in Sichuan,2012(142):24-29.(in Chinese)
[5]钟继茂,李祖宁,吴绍祖,等.福建定点和流动跨断层对台湾强震的反映及特征研究[J].华南地震,2008,28(3):63-70.ZHONG Jimao,LI Zuning,WU Shaozu,et al. Study on responses of Fujian stationary and mobile cross-fault deformations to strong earthquakes in Taiwan[J]. South China Journal of Seismology,2008,28(3):63-70.(in Chinese)
[6]李福利.宁夏跨断层水准测量资料震前异常指标分析[J].防灾减灾学报,2013,29(3):35-38.LI Fuli. Anomalous information analysis of leveling data before the earthquake in Ningxia[J]. Seismological Research of Northeast China,2013,29(3):35-38.(in Chinese)
[7]高宇,曹明明,戴慧芳,等.陕西省榆林地区塌陷地震灾害预测研究[J].自然灾害学报,2014,23(3):213-221.GAO Yu,CAO Mingming,DAI Huifang,et al. Research on prediction of collapse earthquake disaster in Yulin Region,Shanxi Province[J].Journal of Natural Disasters,2014,23(3):213-221.(in Chinese)
[8]王琼,曲延军.投影寻踪自回归模型和应变积累释放模型在新疆地震趋势预测中的应用[J].西北地震学报,2006,28(3):229-234.WANG Qiong,QU Yanjun. Application of project pursuit auto regresson model and strain accumulating and releasing model earthquake trend prediction in Xinjiang[J]. Northwestern Seismological Journal,2006,28(3):229-234.(in Chinese)
[9]蔚赵春.时间序列挖掘算法及其在地震预报中的应用研究[D].上海:上海大学,2005.WEI Zhaochun. Time Series Mining Algorithm and Its Application in Earthquake Prediction[D]. Shanghai:Shanghai University,2005.(in Chinese)
[10]李炜,郑华,邱剑锋,等.基于时间序列相似性匹配算法的地震预测研究[J].四川地震,2010,26(2):4-9.LI Wei,ZHENG Hua,QIU Jianfeng,et al. Earthquake predicting based on amatching a lgorithm of the time series similarity[J]. Earthquake Research in Sichuan,2010,26(2):4-9.(in Chinese)
[11]李英民,董银峰,赖明.地震动瞬时谱估计的Unscented Kalman滤波方法[J].应用数学和力学,2007,28(11):1370-1378.LI Yingmin,DONG Yinfeng,LAI Ming. Instantaneous spectrun estimation of earthquake ground motions based on Unscented Kalman filter Method[J]. Applied Mathematics and Mechanics,2007,28(11):1370-1378.(in Chinese)
[12]陈绍绪,王宝坤,张跃刚.华北地区地倾斜中短期地震前兆识别及预报方法研究[J].地壳形变与地震,1999,19(4):8-14.CHEN Shaoxu,WANG Baokun,ZHANG Yuegang. Idengtification of mediumand short term seimic precursors for ground tilt in north China and study on earthquake prediction methods[J]. Crustal Deformation and Earthquake,1999,19(4):8-14.(in Chinese)
[13]韩晓飞,潘存英,罗词建.基于遗传算法的广义回归神经网络在地震预测中的应用[J].华北地震科学,2012,30(1):48-53.HAN Xiaofei,PAN Cunying,LUO Cijian. Application of generalized regression neural network based on genetic algorithm in earthquake prediction[J]. North China Earthquake Sciences,2012,30(1):48-53.(in Chinese)
[14]赵利飞,王炜.利用神经网络技术对华东地区进行地震预测[J].地震学刊,2002,22(4):21-25.ZHAO Lifei,WANG Wei. The application of neural network technique in the earthquake prediction in East China[J]. Journal of Seismology,2002,22(4):21-25.(in Chinese)
[15]张正禄.工程测量学.第2版[M].武汉:武汉大学出版社,2013.ZHANG Zhenglu. Engineering Surveying. 2nd Edition[M]. Wuhan:Wuhan University Press,2013.(in Chinese)
[16]许国辉,余春林.卡尔曼滤波模型的建立及其在施工变形测量中的应用[J].测绘通报,2004(4):22~23.XU Guohui,YU Chunlin. Establishment of Kalman filtering model and its application in construction deformation measurement[J]. Bulletin of Surveying and Mapping,2004(4):22~23.(in Chinese)
[17]邓自立.卡尔曼滤波与维纳滤波—现代时间序列分析方法[M].哈尔滨:哈尔滨工业大学出版社,2001.DENG Zili. Kalman Filtering and Wiener Filtering-Modern Time Series Analysis Methods[M]. Harbin:Harbin Institute of Technology Press,2001.(in Chinese)
[2] Cheng Wanzheng,Yang Yongling. Deformation rate changes of tectonic belts along boundaries of Yunnan Sichuan block and relation to grouped strong earthquakes[J]. Journal of Geodesy and Geodynamics,2003,23(b12):101-107.
[3]黄卿团,郑韶鹏.福建东南沿海及邻区活动断裂的微地貌研究[J].地球物理学进展,2006,21(4):1099-1107.HUANG Qingtuan,ZHENG Shaopeng. Micro-relief research on active fault in the coast of southeast Fujian and adjacent area[J]. Progress in Geophysics,2006,21(4):1099-1107.(in Chinese)
[4]苏琴,杨永林,向和平.汶川8. 0级地震前后鲜水河断裂带断层活动特征分析[J].四川地震,2012(142):24-29.SU Qin,YANG Yonglin,XIANG Heping. Seismicity character analysis on the Xianshuihe faults before and after the 2008 Wenchuan MS8. 0Earthquake[J]. Earthquake Research in Sichuan,2012(142):24-29.(in Chinese)
[5]钟继茂,李祖宁,吴绍祖,等.福建定点和流动跨断层对台湾强震的反映及特征研究[J].华南地震,2008,28(3):63-70.ZHONG Jimao,LI Zuning,WU Shaozu,et al. Study on responses of Fujian stationary and mobile cross-fault deformations to strong earthquakes in Taiwan[J]. South China Journal of Seismology,2008,28(3):63-70.(in Chinese)
[6]李福利.宁夏跨断层水准测量资料震前异常指标分析[J].防灾减灾学报,2013,29(3):35-38.LI Fuli. Anomalous information analysis of leveling data before the earthquake in Ningxia[J]. Seismological Research of Northeast China,2013,29(3):35-38.(in Chinese)
[7]高宇,曹明明,戴慧芳,等.陕西省榆林地区塌陷地震灾害预测研究[J].自然灾害学报,2014,23(3):213-221.GAO Yu,CAO Mingming,DAI Huifang,et al. Research on prediction of collapse earthquake disaster in Yulin Region,Shanxi Province[J].Journal of Natural Disasters,2014,23(3):213-221.(in Chinese)
[8]王琼,曲延军.投影寻踪自回归模型和应变积累释放模型在新疆地震趋势预测中的应用[J].西北地震学报,2006,28(3):229-234.WANG Qiong,QU Yanjun. Application of project pursuit auto regresson model and strain accumulating and releasing model earthquake trend prediction in Xinjiang[J]. Northwestern Seismological Journal,2006,28(3):229-234.(in Chinese)
[9]蔚赵春.时间序列挖掘算法及其在地震预报中的应用研究[D].上海:上海大学,2005.WEI Zhaochun. Time Series Mining Algorithm and Its Application in Earthquake Prediction[D]. Shanghai:Shanghai University,2005.(in Chinese)
[10]李炜,郑华,邱剑锋,等.基于时间序列相似性匹配算法的地震预测研究[J].四川地震,2010,26(2):4-9.LI Wei,ZHENG Hua,QIU Jianfeng,et al. Earthquake predicting based on amatching a lgorithm of the time series similarity[J]. Earthquake Research in Sichuan,2010,26(2):4-9.(in Chinese)
[11]李英民,董银峰,赖明.地震动瞬时谱估计的Unscented Kalman滤波方法[J].应用数学和力学,2007,28(11):1370-1378.LI Yingmin,DONG Yinfeng,LAI Ming. Instantaneous spectrun estimation of earthquake ground motions based on Unscented Kalman filter Method[J]. Applied Mathematics and Mechanics,2007,28(11):1370-1378.(in Chinese)
[12]陈绍绪,王宝坤,张跃刚.华北地区地倾斜中短期地震前兆识别及预报方法研究[J].地壳形变与地震,1999,19(4):8-14.CHEN Shaoxu,WANG Baokun,ZHANG Yuegang. Idengtification of mediumand short term seimic precursors for ground tilt in north China and study on earthquake prediction methods[J]. Crustal Deformation and Earthquake,1999,19(4):8-14.(in Chinese)
[13]韩晓飞,潘存英,罗词建.基于遗传算法的广义回归神经网络在地震预测中的应用[J].华北地震科学,2012,30(1):48-53.HAN Xiaofei,PAN Cunying,LUO Cijian. Application of generalized regression neural network based on genetic algorithm in earthquake prediction[J]. North China Earthquake Sciences,2012,30(1):48-53.(in Chinese)
[14]赵利飞,王炜.利用神经网络技术对华东地区进行地震预测[J].地震学刊,2002,22(4):21-25.ZHAO Lifei,WANG Wei. The application of neural network technique in the earthquake prediction in East China[J]. Journal of Seismology,2002,22(4):21-25.(in Chinese)
[15]张正禄.工程测量学.第2版[M].武汉:武汉大学出版社,2013.ZHANG Zhenglu. Engineering Surveying. 2nd Edition[M]. Wuhan:Wuhan University Press,2013.(in Chinese)
[16]许国辉,余春林.卡尔曼滤波模型的建立及其在施工变形测量中的应用[J].测绘通报,2004(4):22~23.XU Guohui,YU Chunlin. Establishment of Kalman filtering model and its application in construction deformation measurement[J]. Bulletin of Surveying and Mapping,2004(4):22~23.(in Chinese)
[17]邓自立.卡尔曼滤波与维纳滤波—现代时间序列分析方法[M].哈尔滨:哈尔滨工业大学出版社,2001.DENG Zili. Kalman Filtering and Wiener Filtering-Modern Time Series Analysis Methods[M]. Harbin:Harbin Institute of Technology Press,2001.(in Chinese)
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |