地铁施工安全风险时空耦合机理及实景仿真预警技术研究

Safety control in metro construction is a very complex process. The main reason is that the risks of metro construction are affected by many factors such as site environment, the location of the project, geological conditions, surrounding buildings, underground pipelines, engineering design, construction and engineering methods, on-site workers behaviors and so on. Besides, interactions exist among those factors. Therefore, the risks exist in the metro construction are combined and evolved through the time and space. Through traditional safety control methods, it is difficult to simulate and couple these complex influencing factors. The site works cannot understand which are the dangerous structural components and behaviors, where are they located, what are the affecting areas. In this research, multiple methods including numerical simulation, data mining, Bayesian algorithm and sophisticated network are combined to provide the dynamic analysis of the geotechnical, structural and mechanical properties through the construction progress. Also, based on the four-dimensional model (3D model plus time factor), the spatial and temporal evolvement pattern and risk coupling mechanism are demonstrated by visualization technology. Besides, the dangerous components and its affecting area can be early warned and controlled. A new approach is proposed and it integrates the risk information with a physical, real-world environment whose elements are augmented with the virtual building components. At present, a new era with high-speed rail transit development in China comes into being. This research project not only extends the risk management theory, but also is consistent with the visual simulation technology research and development trends. Therefore, it contains important academic significance and engineering value and it has a broad application prospect.

地铁施工过程中的安全风险受工程地理位置、地质条件、周边环境、工程设计方案、施工工法、现场施工人员和机械行为等多因素综合影响，是随着施工过程在时间和空间上不断发生演化的。现有安全控制手段难以对安全风险的多种影响因素进行耦合模拟，不能有效描述各种风险在时间和空间上的演化规律及其叠加效应，缺乏基于风险耦合分析的危险结构构件、施工作业及其影响区域的可视化预警技术与方法。因此本课题将结合数据挖掘、贝叶斯网络、复杂网络及数值仿真等多种方法，提供随进度变化的岩土、结构、机械多属性力学性能动态分析，基于四维空间（三维空间加时间维）研究施工过程中多种安全风险耦合叠加后的时空演化规律，为现场工人提供一种集成环境感知度的风险叠加实景仿真可视化预警方法。我国正处于地铁工程建设的高峰时期，本研究既适应安全风险控制理论发展的需要，又符合可视化预警技术研究发展趋势，具有十分重要的学术意义和工程价值和应用前景。

本项目结合数据挖掘、贝叶斯网络、复杂网络方法，利用地铁工程施工过程中产生的多源异构的数据，提供随进度变化的岩土、结构、机械多属性力学性能动态分析，并基于四维空间（三维空间加时间维）研究施工过程中多种安全风险耦合叠加后的时空演化规律，从而为现场工人提供一种集成环境感知度的风险叠加实景仿真可视化预警方法。本项目研究解决了现有工程安全控制手段难以对安全风险的多种影响因素进行耦合模拟，不能有效描述各种风险在时间和空间上的演化规律及其叠加效应，缺乏基于风险耦合分析的危险结构构件、施工作业及其影响区域的可视化预警技术与方法等问题。我国正处于地铁工程建设的高峰时期，本项目研究所取得的进展和成果能够有效地对地铁施工过程中的安全风险进行预判和预警，从而规避重大风险事件和事故的发生。其取得的研究成果既适应安全风险控制理论发展的需要，又符合可视化预警技术研究发展趋势，具有十分重要的学术意义和工程应用前景。