A Point Cloud Virtual Trial Assembly Method for Preventive Quality Control of Prefabricated Bridge

Abstract

With the development of bridge construction equipment and technologies, prefabricated girders have become larger and heavier, posing an increasingly higher challenge to the trial assembly quality control. Hence, more attention has been devoted to the virtual trial assembly (VTA) due to the success of laser techniques in bridge construction. Based on the Shenzhen–Zhongshan link, a world-class sea-crossing project, we develop a laser point cloud-driven VTA method for prefabricated girder preventive quality control, addressing challenges arising from the large volume, complex interface configuration, and diverse attachment noises. Specifically, the interface matching feature of the target bridge girder is first extracted from the raw 3D laser point cloud model. Then, the pose adjustment procedure is conducted for a precise virtual assembly, followed by the proposed assembly criteria, namely, the degree of matching (DoM), for quantitative evaluation. Our proposed VTA solution has been successfully applied in the whole construction process of the Shenzhen–Zhongshan link; besides, a comprehensive numerical study has also been conducted, inspired by this project, to prove the accuracy of our proposed VTA method. Results demonstrate that the extracted interface dimensional features are precise with error controlled under 1%. Then, assembly results between two interfaces show that 95% points of the interfaces can remain under 0.005 m in terms of DoM values, demonstrating the performance of the proposed method.