基于贝叶斯网络的缓坡场地震液化侧移评估——以台湾集集地震为例  

Assessment of seismic liquefaction-induced lateral spreading in gentle slope field based on Bayesian network:a case of the Chi-chi earthquake,Taiwan

作　　者：张政 胡记磊 刘华北[1] ZHANG Zheng;HU Jilei;LIU Huabei(School of Civil Engineering and Mechanics,Huazhong University of Science and Technology,Wuhan 430074,China)

机构地区：[1]华中科技大学土木工程与力学学院,湖北武汉430074

出　　处：《自然灾害学报》2018年第6期127-132,共6页Journal of Natural Disasters

基　　金：国家自然科学基金青年科学基金项目(41702303)~~

摘　　要：本文基于贝叶斯网络方法,综合考虑了地震液化侧移的13个影响因素及其不确定性,建立了适用于缓坡场地的地震液化侧移贝叶斯网络评估模型。通过我国台湾集集地震液化灾害的算例分析,验证了地震液化侧移的贝叶斯网络评估模型的准确可靠性,并通过分析模型性能的评估指标和人工神经网络模型进行了对比。本文建立的地震液化侧移贝叶斯网络评估模型不仅可以预测场地是否会发生液化,而且可以在此基础上进一步评估场地液化引起的侧移量等级。Based on the Bayesian network method,this paper synthetically considers 13 influencing factors and uncertainty of seismic liquefaction lateral spreading,and establishes a Bayesian network assessment model for seismic liquefaction lateral spreading for gentle slope field.Through the example analysis of the liquefaction disasters in Chichi earthquake,Taiwan,China,the accuracy and reliability of the Bayesian network assessment model for seismic liquefaction lateral spreading are illustrated,and its evaluation indices for model performance are compared with the artificial neural network model.The Bayesian network assessment model established in this paper not only can predict liquefaction potential,but also can further assess the level of liquefaction-induced lateral spreading.

关 键 词：贝叶斯网络 地震液化 侧移 缓坡场地 评估 

分 类 号：X43[环境科学与工程—灾害防治] X9