建筑幕墙全生命周期智能化：数字孪生、智能建造与云边端协同

Building Facade Full-Life-Cycle Intelligence: Digital Twin, Intelligent Construction, and Cloud-Edge-Device Collaboration

摘要: 建筑幕墙在轻量化、抗震安全、设计灵活性、施工效率、节能环保等多个维度具有显著优势，这些特性使其成为现代高层建筑外围护结构的主导选择，但其长期面临设计效率低、施工精度差、运维风险高三重困境。人工智能与数字孪生技术的引入，正在推动幕墙系统从“单点应用”走向“系统集成”。文章系统梳理了AI与数字孪生在建筑幕墙领域的最新研究与应用进展，从感知、决策、执行三个维度构建了幕墙智能体的技术框架；同时围绕设计施工、智能建造、运维监测三大环节展开梳理，讨论了数据标准缺失、模型泛化能力不足等当前瓶颈，并对物理信息神经网络、多模态大模型、自主智能体系统及生成式AI辅助设计等未来发展方向进行展望，旨在为建筑智能化领域的研究者与工程实践者提供系统性的理论参考。

Abstract: Building facades offer significant advantages in multiple dimensions, including light weight, seismic safety, design flexibility, construction efficiency, and energy conservation. These characteristics have made them the dominant choice for the exterior enclosure of modern high-rise buildings. However, the facade industry has long been plagued by three persistent challenges: low design efficiency, poor construction accuracy, and high operational risks. The introduction of artificial intelligence and digital twin technologies is driving facade systems from “standalone applications” toward “system integration”. This paper systematically reviews the latest research and application progress of AI and digital twins in the field of building facades, and constructs a technical framework for the facade agent from three dimensions: perception, decision-making, and execution. The review is organized around the stages of design and construction, intelligent construction, and operation and maintenance. It discusses current bottlenecks such as the lack of data standards and insufficient model generalization, and looks ahead to future directions including physics-informed neural networks, multimodal foundation models, autonomous agent systems, and generative AI-assisted design. The aim is to provide researchers and engineering practitioners in the field of building intelligence with a systematic theoretical reference.