城市地铁施工安全风险云的分布演化机理与控制方法研究

The safety risk cloud of urban metro construction is a comprehensive reflection of safety status and spatio-temporal distribution of metro construction system. It is emerging, changing with the construction process on the spatial and temporal scales until the endless of disaster. However, there is the lack of awareness and research on it both at home and abroad. This paper focuses on the systematical study on the distribution and evolution mechanism and controlling for safety risk cloud in urban metro construction under the guide of the complex network theory, data-driven method and decoupling control theory, together with the research methods of theoretical modeling, data mining, computational experiment, field verification and so on. The spatial and temporal distribution and difference characteristics of safety risk in metro construction are analyzed in this project. The definition of the safety risk cloud in metro construction is clarified. The influence factors and heterogeneous coupling mechanism of the safety risk cloud in metro construction is studied thoroughly. The multi-agent-based complex network model for the safety risk cloud in metro construction is proposed. With this model, the spatio-temporal dynamic effects and evolution law of disaster of the safety risk cloud in metro construction are under covered. The probability distribution estimation method based on data-driven theory for the safety risk cloud in metro construction is proposed. And the dynamic control strategy based on the decoupling control theory for the safety risk cloud in metro construction is proposed. The results of the research provide the theoretical basis and methodological support for the safety risk analysis and decision in metro construction, and it is of important theoretical and practical significance to improve the safety management of urban construction in China.

城市地铁施工安全风险云是城市地铁施工系统安全状态及其时空分布情况的综合反映，并随着施工过程在时空尺度上不断涌现、演化直至灾变，但国内外对此缺乏系统认识和研究。本课题结合复杂网络、数据驱动、解耦控制等理论方法，采用理论建模、数据挖掘、计算实验、工程验证等多种研究手段，重点围绕城市地铁施工安全风险云的分布演化机理与控制方法这一主题开展系统性研究。通过分析地铁施工安全风险的时空分布特征及时空差异特性，明确地铁施工安全风险云的内涵界定，深入研究地铁施工安全风险云的影响因素及其异质耦合机理；建立基于多主体复杂网络的地铁施工安全风险云模型，揭示地铁施工安全风险云的时空动力效应与灾变演化规律；构建基于数据驱动的地铁施工安全风险云概率分布估计方法，提出基于解耦原理的地铁施工安全风险动态控制策略，为城市地铁施工安全风险分析决策提供理论依据和方法支持，对提高我国城市地铁施工安全管理水平具有重要理论和实践意义。

我国地铁建设步入快速发展时期，建设规模和速度已居世界之首。但是，地铁施工面临复杂的地质水文条件，不可避免需要穿越城市内部大量敏感的城建设施和生命线系统，是高风险的人-机-环复杂系统工程。特别是城市地铁施工安全事故同时具有生产安全事故和公共安全事故双重特性，一旦发生将导致重大人员和财产损失，造成的社会影响严重深远。面向国家经济社会发展与工程实践的重大需求，本项目重点围绕地铁施工安全中的“人-机-环境”风险云耦合机理与演化规律这一科学问题，构建城市地铁施工安全风险云的复杂网络模型和分析方法，揭示城市地铁施工安全风险云的异质耦合机理、时空动力效应及灾变演化规律，建立了数据驱动的地铁施工安全风险预警方法及系统，研发了地铁施工安全控制物联网体系架构及泛在感控设备，获得2018年教育部科技进步一等奖，2017年香港建造业议会国际创新大奖（唯一最高奖项），第十一届湖北省专利银奖，同时还获得国家发明专利授权2项，在SAFETY SCIENCE、AUTCON、RESS、ASCE COMPUT等本领域国际顶级期刊上发表学术论文10余篇，研究成果分别在香港、深圳、沈阳、武汉等多座城市地铁工程中成功应用，并推广到其他土木与建筑工程领域中，有力保障了全国首条长江地铁隧道、国内最大直径公铁合建隧道等重大工程的施工安全，丰富和促进了我国地铁工程安全风险控制实践。