基于BIM和多源数据集成的地铁施工精准风险评估与实时控制

The tight construction period and the frequent occurrence of risk accidents in China's underground tunnel construction have caused great socio-economic losses, which highlight the specialty and importance of risk management on underground tunnel constructions, and the urgent need to improve risk management and control. From the perspective of uncertainty artificial intelligence and based on the BIM technology, this study adopts the “reverse thinking” to identify risks, so as to avoid any omissions.①The forward analysis "analyses uncertainty based on certainty", where 4M1E is used to integrate the quality factors and risk management. Based on the data fusion theory, the cloud model, the DS evidence theory and the matter-element extension theory, a new theory of precise risk assessment is forwarded. According to the catastrophe theory, the catastrophe progression method is used to study the time series characteristics of complex underground tunnel construction systems, so as to reveal the law of risk evolution and to warn risks dynamically. To achieve a real-time control of risks and provide a new tool for decision making at the presence of underground tunnel construction risks, an integrated approach, which includes management integration, data fusion and technology integration, is adopted.②The reverse analysis, based on the failure knowledge management, uses ontology technology to build a failure knowledge database and supplemental rules to control risk factors, so as to build a new pattern of closed-loop risk identification modal. These methods are implemented and verified during the construction of Xiamen underground tunnel Line 3 and Line 2 projects. By studying precise risk assessment and real-time control, this study will help reduce the risk of underground tunnel constructions in our country, which has theoretical and application significance.

我国地铁建设规模大工期紧，风险事故频发，造成极大的社会经济损失，凸显了地铁施工风险管理特殊性和重要性，亟待提高风险管理和控制水平。从不确定性人工智能的视角，基于BIM技术，逆向思维风险识别，以避免风险遗漏。①在正向分析中，“以确定性分析不确定性”，采用4M1E整合质量因素和风险管理；并基于数据融合理论，综合运用云模型、D-S证据理论和物元可拓理论，提出精准风险评估新理论。依据突变理论，采用突变级数法研究地铁施工复杂系统的时序特征，揭示风险演变规律，动态预警风险；采用集成方法，进行管理整合、数据融合和技术集成，实现风险实时控制，提供地铁施工风险决策支持新工具。②在逆向分析中，基于失败知识管理，采用本体技术构建失败知识库和风险因素补充规则，构建风险识别闭环新模式。通过厦门地铁三号线和二号线工程进行实践验证。本项目的研究精准风险评估和实时控制，有助于降低我国地铁施工风险，具有理论意义和应用价值。

我国地铁建设规模大工期紧，风险事故频发，凸显了地铁施工风险管理特殊性和重要性，亟待提高风险管理和控制水平。本项目以数据融合理论、失败学、控制论等多学科理论为基础，着眼于我国地铁施工风险的现实需要，力求克服风险管理的不完全、主观性、不准确、不直观和不及时的难题。.本课题①提出了“以确定性分析不确定性”的风险分析新思路，采用4M1E整合质量与风险管理，统一风险评价指标体系；采用逆向思维，从正反双向分析（before-and-after）两个方向进行风险因素识别，“完全”识别风险，提出风险因素补充规则，实现风险识别闭环新模式。②依据云理论、证据融合理论和物元理论，建立融合评估模型和定量评估方法，实现了施工风险多源数据融合，提出精准风险评估新理论。③针对风险预警均为静态结果，本课题提出了基于云证据理论的施工风险预警方法，实现了不确定多源信息的融合和定性与定量指标的转换，实现了提前2天进行风险预警。④采用BIM实现施工风险可视化表达，同时基于BIM和4M1E整合质量和风险管理，存储施工质量和风险的多源数据信息。以集成方法为主线，进行管理整合、数据融合、技术集成的多层次、多维度和多方位集成，采用BIM与IoT技术集成设计多源数据采集系统，设计精准评估和实时控制整体方案，提供地铁施工风险的实时控制新工具。⑤并进一步深入探讨风险对策智能生成领域，从纯文本的本体技术的知识驱动的路线与方法到数据驱动的深度学习与NLP技术的结合，再到跨模态的多源数据融合的智能生成。⑥最后，以厦门地铁三号线过海通道和二号线作为工程实证，进行了理论与方法验证和工程应用。.本课题实现了定量评估、动态预警、实时控制和施工风险可视化，从而将不确定、不可呈现的施工风险变为可以准确定量、精确定位、动态预警、实时可控。上述研究与风险对策智能生成，不仅是施工风险精准评估与实时控制，而且对施工风险管理的理论具有创新意义。

内容获取失败，请点击重试