魏伟青，陈磊，骆汉宾

摘要:越江地铁隧道在建设过程中需要克服岩土体强度和围岩稳定性低、地下水压力大、局部埋深浅等不利条件，安全风险极大。聚类分析将越江隧道所处工程地质分为若干类别，利用数据挖掘发现隐藏在地质数据中的内在规律。本文针对越江地铁隧道地质特点，构建了地质安全评价指标体系，并以武汉地铁四号线越江隧道为例，通过 K-means 快速聚类方法对比不同 K 值和不同评价指标设定的多组聚类结果，与不同类型盾构管片的实际分布情况进行比较，验证了该方法的可行性和结果的合理性。

关键词: 越江隧道; 地质环境; 安全评价; 聚类分析

GeologicalSafety Evaluation of Cross-river Metro Tunnel Based on Clustering Analysis

Abstract: Cross-river metrotunnel construction must overcome low rock mass strength and the stability ofsurrounding rock， strong ground waterpressure， the local embeddedshallow depth and other unfavorable conditions， which were with high safety risk． Cross-river metro tunnel can bedivided into several categories， according to clusteranalysis． As an importantcontent of data mining， it can find outinherent law hiding in the geological data．This paper was based on geological feature of cross-river metro tunnel， and K-means quick clustering wasproposed to clear and definite the evaluation process and method． Then the cross-river metrotunnel in Wuhan metro line 4 was selected as an engineering example． By comparing different values ofK and multiple sets of clustering results with different evaluation index， and matching with the actualdistribution of different types of shield segments， the feasibility of this method， the rationality of the resultsis verified．

Keywords: cross-rivertunnel; geological environment; safety evaluation; cluster analysis

下载地址：

http://www.cnki.net/KCMS/detail/detail.aspx?QueryID=4&CurRec=1&recid=&filename=WHCJ201501013&dbname=CJFDTEMP&dbcode=CJFQ&pr=&urlid=&yx=&uid=WEEvREcwSlJHSldTTGJhYlJRYk56K3B1V3BtTmh0YnVFWnh5dHFsUjZkYzdGNVBTSGRmTmRvVTNGZmw2ZiswVGlJOD0=$9A4hF_YAuvQ5obgVAqNKPCYcEjKensW4IQMovwHtwkF4VYPoHbKxJw!!&v=MTkxNjlEaDFUM3FUcldNMUZyQ1VSTCtmWWVkdUZ5L25WN3JMTWlYSVpMRzRIOVRNcm85RVo0UjhlWDFMdXhZUzc=