复杂环境地铁隧道数字孪生系统的研究
Research on Digital Twin System for Complex Environment Subway Tunnels

为降低地铁隧道因各项病害而产生的结构安全性风险，本研究以数字孪生技术为基础，构建地铁隧道的智能化管理系统平台。通过设计隧道无人巡检车，实现对隧道的智能巡检，并通过大数据分析及时获取和处理隧道的病害与形变信息。结合神经网络和机器学习的方法，采用Transformer模型实现对隧道沉降的精准预测。在综合考虑地铁隧道环境智能重构、数据互联、感知交互、业务协作等共性支撑技术的基础上，本研究构建了地铁隧道全节点、多维度实景数字孪生系统。该系统能够实现隧道的智能巡检、故障诊断、沉降预测、智能交互和三维展示，从而显著提高地铁隧道的安全管理和设备维护效率。
To mitigate the structural safety risks arising from various ailments in subway tunnels, this study leverages digital twin technology to establish an intelligent management system platform for subway tunnels. Through the design of an unmanned tunnel inspection vehicle, intelligent inspections of the tunnel are achieved. The study utilizes large-scale data analysis to promptly acquire and process information related to defects and deformations within the tunnel. By employing neural networks and machine learning methodologies, the Transformer model is utilized to achieve precise predictions of subsidence within the tunnel. Integrating various common supporting technologies, such as intelligent reconstruction of the subway tunnel environment, seamless data interconnection, perceptual interaction, and collaborative business strategies, this research establishes a subway tunnel full-node, multi-dimensional real-scene digital twin system. This system facilitates intelligent inspections, fault diagnosis, accurate subsidence predictions, intelligent interaction, and three-dimensional displays within the tunnel. Consequently, it significantly enhances the safety management and equipment maintenance efficiency of subway tunnels.