Xinyi Yu

The manual splicing of steel arches in Tunnel Boring Machine construction is notoriously inefficient, hazardous, and quality-inconsistent, posing a significant bottleneck to tunneling automation. To address this challenge, this paper proposes a novel optimization framework for the motion/force transmission performance of a closed-chain robotic end-effector specifically designed for automated steel arch splicing. Based on screw theory, a comprehensive transmission performance model is established and the Global Transmission Index is employed as a key metric to optimize the transmission angle range, thereby ensuring both high efficiency and stability during the grasping process. The proposed method integrates actual spatial constraints and grasping requirements of steel arches into a systematic design optimization process. Experimental results demonstrate that the optimized end-effector completes the splicing task in only 90 seconds, which is more than three times faster than traditional manual splicing, while achieving reliable grasping and a docking accuracy of \(\pm\)3 mm. This study provides both theoretical and practical advancements in robotic tunnel support, offering a viable solution for enhancing the intelligence and safety of TBM construction.