基于数据挖掘的砂土地震液化预测模型研究
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
在对砂土地震液化各影响因素分析的基础上,引入整合式模糊聚类神经网络的数据挖掘模型。该模型以模糊系统为框架,结合聚类分析技术和神经网络模式,建立砂土液化与其影响因素之间的非线性关系。运用所建立的模型,对国内外400组砂土地震液化资料进行分析,来预测砂土液化现象。数据挖掘结果的误判率及分类矩阵表明,文章的系统模式在判断砂土地震液化时能取得较好的效果。
An integral Fuzzy Clustering Neural Network model is introduced based on the analysis of the earthquake induced sand liquefaction factors.The model which is in the framework of fuzzy system,combined with clustering analysis and neural network,established a nonlinear relationship between liquefaction and factors.In order to predict the liquefaction phenomenon,400 groups of liquefaction data from domestic and abroad was analyzed.The misjudgement ratio and classification matrix of data mining results indicate that the systematic model in this paper can effectively predict earthquake induced sand liquefaction.
引文
[1]罗国煜,陈新民,李晓昭,等.城市环境岩土工程[M].南京:南京大学出版社,2000.
    [2]马骥.基于MATLAB的BP神经网络在砂土液化评价中的应用[D].南京:南京大学硕士论文.2002.
    [3]翁焕学.砂土地震液化模糊综合评判实用方法[J].岩土工程学报.1993,15(2):74-79.
    [4]Juang C H,Chen C J.CPT-Based liquefaction evaluation usingartificial neural networks[J].Computer-Aided Civil andInfrastructure Engineering.1999,14(3):221-229.
    [5]陈新民,罗国煜.地震砂土液化可能性的非确定性灰色预测方法[J].桂林工学院学报.1997,17(2):106-109.
    [6]Jiawei Han,Micheline Kamber.Data mining:Concepts andtechniques[M].Morgan Kaufmann.2001.
    [7]廖少明,刘朝明,王建华,等.地铁深基坑变形数据的挖掘分析与风险识别[J].岩土工程学报.2006.28(S0):1897-1901.
    [8]陈季,李煜仕,张延任,等.以整合式模糊群聚类神经网路评估土壤液化[J].中国土木水利工程学刊(台湾).2005,17(2):335-347.
    [9]Chern S G,Hu R F,Chang Y J,Tsai I F.Fuzz-ART neuralnetworks for predicting Chi-Chi earthquake inducedliquefaction in Yuan-Lin area[J].Journal of Marine Scienceand Technology.2002,10(1):21-32.
    [10]张乾飞,徐洪钟,吴中如,等.基于模糊聚类的神经网络模型及其在渗流分析中的应用[J].水力发电学报.2002,2:37-43.
    [11]曲永玲,陈新民.聚类分析法在土层划分中的应用[J].南京工业大学学报.2005,27(5):85-88.

版权所有:© 2023 中国地质图书馆 中国地质调查局地学文献中心