大型复杂龙骨-玻璃面板结构建造过程中,龙骨弯曲且扭转,建成质量较难测量,影响面板安装；此外,传统玻璃面板深化过程存在效率不高,智能化程度低等缺点。点云和启发式算法为骨架竣工信息获取和面板智能深化设计带来了可行的解决方案。为此,以广州南沙国际金融论坛会址为工程背景,开展基于点云数据和启发式算法的大型复杂龙骨-面板结构智能建造技术研究：针对采光顶深化设计需求,提出适用于龙骨的逆向建模方法；针对骨架几何描述困难问题,提出多边形弯扭构件轴线提取算法；针对手动划分面板效率低等问题,提出智能化的单元提取方法；基于Guillotine算法,采用不同策略生成面板排布方案并进行评估。研究结果表明,所提逆向建模方法能准确获取龙骨竣工模型,面板单元智能划分方法简单有效,基于启发式算法进行面板排布能快速获得合理下料方案。研究成果可供同类结构数字化建造转型参考。

In the construction of large and complex frame-panel structures, it is difficult to measure the quality of the frame due to bending and torsion, which affects the panel installation. In addition, the traditional glass panel deepening design suffers from low efficiency and poor intelligence. Point cloud data (PCD) and heuristic algorithms provide a perfect solution for the status obtain of completed frame and intelligent of panel design. According to the Buildings of International Financial Forum located in Nansha District, Guangzhou city, the intelligent construction techniques for large-scale complex frame- panel structures based on PCD and heuristic algorithms were studied in this paper: Conversational modeling methods suitable for frames have been proposed in response to the design requirements of panels; The algorithms was proposed to extract the axis of the curved and twisted pip which have a polygon section to address the difficulty in skeleton geometry description; A unit-wise automatic extraction method has been proposed to improve the efficiency of panel partitioning; different strategies were used to generate panel layout schemes based on the Guillotine algorithm, and layout results were evaluated. The research show that the reverse modeling method can accurately obtain the frame model, the intelligent partitioning method for panel is simple and effective, and excellent cutting plans can be quickly obtained using heuristic algorithms. The results can be used as a reference for digital construction transformations of similar structures.

TU398

高层钢-混凝土混合结构的智能建造算法研究