面向大数据的城市地下工程施工期安全风险评估方法研究

项目名称： 面向大数据的城市地下工程施工期安全风险评估方法研究

项目编号： No.41472288

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 地质学

项目作者： 王浩

作者单位： 中国科学院武汉岩土力学研究所

项目金额： 90万元

中文摘要： 大数据时代即将到来，大规模、全方位、多场的城市地下工程监测将成为现实。大数据环境下的风险评估有何不同？相关的研究工作很少见。本项目依托城市轨道交通第三方监测工程，探索将大数据的有关技术及方法完整地实施于监测数据分析和风险评估中，形成示范性项目和成套技术供相关工程借鉴，提升地下工程风险评估的理论水平。首先，针对今后大规模的数据采集、存储需求，研究大数据的分布式存储技术、自动化监测传感器的布置策略、组网方式、数据存储技术和网络通信可靠性；其次，应用大数据技术中的数据挖掘方法来快速识别监测异常和判断成因：依靠数据库表的计算来辨识异常和风险源，打破传统监测异常分析中对曲线形态判别的依赖；基于数据库表进行监测数据与多个影响因素之间的关联分析和成因分析；最后，在对监测数据、巡视检查文字和视频图像进行数据挖掘的基础上，结合常规风险评估理论，形成大数据多源融合环境下的城市地下工程综合风险评估方法。

中文关键词： 地下工程；大数据；风险评估；安全监测；数据挖掘

英文摘要： The era of Big Data is coming, and large-scale, all-round, multi- field monitoring of urban underground engineering will become a reality. The risk assessment methodology under big data environment is rarely reported. Relying on a third-party monitoring of urban rail transit project, this research will adopt big data relevant technologies to implement a complete monitoring data analysis and risk assessment for an urban underground engineering project. Then, a demonstration model and associated comprehensive technical references will be prepared to guide related projects, thus the theoretical level of risk assessment for underground engineering may be improved. First, to meet the demand of large-scale data acquisitions and storage requirements, distributed storage technology of big data will be employed to examine the layout strategy, networking, data storage and network communication reliability for large-scale automated monitoring sensors. Second, several data mining methods of big data technology will be applied to quickly identify and determine the causes of abnormal in monitoring results. For example, to identify anomalies and risk sources through the calculation of the database tables instead of the traditional method of curve shape discrimination; to quick conduct cause analysis, result interpretation and association analysis between monitoring data and multiple influence factors based on a database tables. Finally, combining with the commonly used risk assessment theories, a comprehensive multi-source data fusion risk assessment method for urban underground engineering under big data environment will be formed based on data mining from monitoring data, texts of field inspection and video images.

英文关键词： underground engineering;big data;risk assessment;safty monitoring;data mining