基于贝叶斯网络的自由场地震液化沉降评估
|
摘要
基于贝叶斯网络方法,综合考虑地震参数、土体参数和场地条件等12个影响因素,结合场地的液化势和液化潜能指数,建立了地震液化沉降的贝叶斯网络评估模型。通过算例分析,与径向基神经网络方法和I&Y简化计算方法的评估性能对比,发现地震液化沉降的贝叶斯网络评估模型的优势明显;该模型不仅有较好的评估精度和可靠性,而且还可以进行逆向因果推理。对两个机器学习模型进行敏感因素分析发现,在12个影响因素中,地表峰值加速度、地震持续时间和标准贯入锤击数为较敏感因素,和I&Y简化算法考虑的参数基本一致。
Based on the Bayesian network method,a Bayesian network model for assessing seismic liquefactioninduced settlement was constructed,in which 12 significant factors including earthquake parameters,soil parameters and field conditions combining with the liquefaction potential and liquefaction potential index were considered. Through some cases study,it is shown the Bayesian network model has obvious advantages in the assessment performance,comparing with the RBF( Radial Basis Function) neural network method and I & Y( Ishihara & Yoshimine) simplified calculation method. The Bayesian network model not only has better assessment accuracy and reliability,but can also perform reverse causal reasoning. In the analysis of sensitive factors to the two machine learning models,the ground peak acceleration,duration of earthquake and standard penetration test blow count are more sensitive among the 12 factors,which are the same as those considered in the I & Y simplified calculation method.
Based on the Bayesian network method,a Bayesian network model for assessing seismic liquefactioninduced settlement was constructed,in which 12 significant factors including earthquake parameters,soil parameters and field conditions combining with the liquefaction potential and liquefaction potential index were considered. Through some cases study,it is shown the Bayesian network model has obvious advantages in the assessment performance,comparing with the RBF( Radial Basis Function) neural network method and I & Y( Ishihara & Yoshimine) simplified calculation method. The Bayesian network model not only has better assessment accuracy and reliability,but can also perform reverse causal reasoning. In the analysis of sensitive factors to the two machine learning models,the ground peak acceleration,duration of earthquake and standard penetration test blow count are more sensitive among the 12 factors,which are the same as those considered in the I & Y simplified calculation method.
引文
[1]KAWASAMI H.General report on the Niigata earthquake of1964[M].Tokyo:Tokyo Electrical Engineering College Press,1968.
[2]LEE D H,JUANG C H,KU C S.Liquefaction performance of soils at the site of a partially completed ground improvement project during the 1999 Chi-Chi earthquake in Taiwan[J].Canadian Geotechnical Journal,2001,38(6):1241-1253.
[3]COX B R,BOULANGER R W,TOKIMATSU K,et al.Liquefaction at strong motion stations and in Urayasu city during the 2011 Tohoku-Oki earthquake[J].Earthquake Spectra,2013,29(1):55-80.
[4]TOKIMATSU K,SEED H B.Evaluation of settlements in sands due to earthquake shaking[J].Journal of Geotechnica Engineering,ASCE,1986,113(8):864-878.
[5]ISHIHARA K,YOSHIMINE M.Evaluation of settlements in sand deposits following liquefaction during earthquake[J].Soils and Foundations,1992,32(1):173-188.
[6]叶斌,叶冠林,长屋淳一.砂土地基地震液化沉降的两种简易计算方法的对比分析[J].岩土工程学报,2010,(增刊2):33-36.YE Bin,YE Guanlin,NAGAYA Junichi.Comparison of two simple methods for assessing subsidence of sandy ground caused by liquefaction in earthquake[J].Chinese Journal o Geotechnical Engineering,2010,(Sup 2):33-36.
[7]CETIN K O,BILGE H T,WU J,et al.Probabilistic mode for the assessment of cyclically induced reconsolidation(volumetric)settlements[J].Journal of Geotechnical and Geoenvironmental Engineering,2009,135(3):387-398.
[8]陈国兴,李方明.基于RBF神经网络模型的砂土液化震陷预估法[J].自然灾害学报,2008,17(1):180-185.CHEN Guoxing,LI Fangming.Seismic settlement estimation of sand liquefaction based on RBF neural network model[J].Journal of Natural Disasters,2008,17(1):180-185.
[9]郭小东,田杰,王威,等.基于GA-SVR的建筑物液化震陷预测方法[J].北京工业大学学报,2011,37(6):829-835.GUO Xiaodong,TIAN Jie,WANG Wei,et al.Method for building settlements prediction due to earthquake liquefaction based on GA-SVR[J].Journal of Beijing University o Technology,2011,37(6):829-835.
[10]BAYRAKTARLI Y Y.Application of Bayesian probabilistic networks for liquefaction of soil[C]∥6th International Ph DSymposium in Civil Engineering.Zurich:Institute o Structural Engineering,2006.
[11]HUANG H W,ZHANG J,ZHANG L M.Bayesian network for characterizing model uncertainty of liquefaction potentia evaluation models[J].KSCE Journal of Civil Engineering,2012,16(5):714-722.
[12]HU Jilei,TANG Xiaowei,QIU Jiangnan.Assessment of Seismic liquefaction potential based on Bayesian network constructed from domain knowledge and history data[J].Soil Dynamics and Earthquake Engineering,2016,89:49-60.
[13]徐国祥.统计预测和决策[M].上海:上海财经大学出版社,2012.
[14]张连文,郭海鹏.贝叶斯网引论[M].北京:科学出版社,2006.
[15]ZHANG G,ROBERTSON P K,BRACHMAN R W I.Estimating liquefaction-induced ground settlements from CPTfor level ground[J].Canadian Geotechnical Journal,2002,39(5):1168-1180.
[16]IWASAKI T,TOKIDA K,TATSUOKA F,et al.Microzonation for soil liquefaction potential using simplified methods[C]∥Proc.3rd International Earthquake Microzonation Conference.Seattle:National Science Foundation,1982.
[17]HWANG J H,YANG C W.Verification of critical cyclic strength curve by Taiwan Chi-Chi earthquake data[J].Soil Dynamics and Earthquake Engineering,2001,21(3):237-257.
[18]CETIN K O,SEED R B,KIUREGHIAN A D,et al.SPT-based probabilistic and deterministic assessment of seismic soil liquefaction initiation hazards:PEER-2000/05[R].Berkeley:Pacific Earthquake Engineering Research,2000.
[19]BRIER G W.Verification of forecasts expressed in terms of probability[J].Monthly Weather Review,1950,78(1):1-3.
[20]UENG T S,WU C W,CHENG H W,et al.Settlements of satruated clean sand deposits in shaking table test[J].Soil Dynamics and Earthquake Engineering,2010,30(1/2):50-60.
[2]LEE D H,JUANG C H,KU C S.Liquefaction performance of soils at the site of a partially completed ground improvement project during the 1999 Chi-Chi earthquake in Taiwan[J].Canadian Geotechnical Journal,2001,38(6):1241-1253.
[3]COX B R,BOULANGER R W,TOKIMATSU K,et al.Liquefaction at strong motion stations and in Urayasu city during the 2011 Tohoku-Oki earthquake[J].Earthquake Spectra,2013,29(1):55-80.
[4]TOKIMATSU K,SEED H B.Evaluation of settlements in sands due to earthquake shaking[J].Journal of Geotechnica Engineering,ASCE,1986,113(8):864-878.
[5]ISHIHARA K,YOSHIMINE M.Evaluation of settlements in sand deposits following liquefaction during earthquake[J].Soils and Foundations,1992,32(1):173-188.
[6]叶斌,叶冠林,长屋淳一.砂土地基地震液化沉降的两种简易计算方法的对比分析[J].岩土工程学报,2010,(增刊2):33-36.YE Bin,YE Guanlin,NAGAYA Junichi.Comparison of two simple methods for assessing subsidence of sandy ground caused by liquefaction in earthquake[J].Chinese Journal o Geotechnical Engineering,2010,(Sup 2):33-36.
[7]CETIN K O,BILGE H T,WU J,et al.Probabilistic mode for the assessment of cyclically induced reconsolidation(volumetric)settlements[J].Journal of Geotechnical and Geoenvironmental Engineering,2009,135(3):387-398.
[8]陈国兴,李方明.基于RBF神经网络模型的砂土液化震陷预估法[J].自然灾害学报,2008,17(1):180-185.CHEN Guoxing,LI Fangming.Seismic settlement estimation of sand liquefaction based on RBF neural network model[J].Journal of Natural Disasters,2008,17(1):180-185.
[9]郭小东,田杰,王威,等.基于GA-SVR的建筑物液化震陷预测方法[J].北京工业大学学报,2011,37(6):829-835.GUO Xiaodong,TIAN Jie,WANG Wei,et al.Method for building settlements prediction due to earthquake liquefaction based on GA-SVR[J].Journal of Beijing University o Technology,2011,37(6):829-835.
[10]BAYRAKTARLI Y Y.Application of Bayesian probabilistic networks for liquefaction of soil[C]∥6th International Ph DSymposium in Civil Engineering.Zurich:Institute o Structural Engineering,2006.
[11]HUANG H W,ZHANG J,ZHANG L M.Bayesian network for characterizing model uncertainty of liquefaction potentia evaluation models[J].KSCE Journal of Civil Engineering,2012,16(5):714-722.
[12]HU Jilei,TANG Xiaowei,QIU Jiangnan.Assessment of Seismic liquefaction potential based on Bayesian network constructed from domain knowledge and history data[J].Soil Dynamics and Earthquake Engineering,2016,89:49-60.
[13]徐国祥.统计预测和决策[M].上海:上海财经大学出版社,2012.
[14]张连文,郭海鹏.贝叶斯网引论[M].北京:科学出版社,2006.
[15]ZHANG G,ROBERTSON P K,BRACHMAN R W I.Estimating liquefaction-induced ground settlements from CPTfor level ground[J].Canadian Geotechnical Journal,2002,39(5):1168-1180.
[16]IWASAKI T,TOKIDA K,TATSUOKA F,et al.Microzonation for soil liquefaction potential using simplified methods[C]∥Proc.3rd International Earthquake Microzonation Conference.Seattle:National Science Foundation,1982.
[17]HWANG J H,YANG C W.Verification of critical cyclic strength curve by Taiwan Chi-Chi earthquake data[J].Soil Dynamics and Earthquake Engineering,2001,21(3):237-257.
[18]CETIN K O,SEED R B,KIUREGHIAN A D,et al.SPT-based probabilistic and deterministic assessment of seismic soil liquefaction initiation hazards:PEER-2000/05[R].Berkeley:Pacific Earthquake Engineering Research,2000.
[19]BRIER G W.Verification of forecasts expressed in terms of probability[J].Monthly Weather Review,1950,78(1):1-3.
[20]UENG T S,WU C W,CHENG H W,et al.Settlements of satruated clean sand deposits in shaking table test[J].Soil Dynamics and Earthquake Engineering,2010,30(1/2):50-60.