跑道
机器人
骨科手术
容错
计算机科学
控制理论(社会学)
人工智能
外科
医学
分布式计算
控制(管理)
历史
考古
作者
Chen Zhao,Tianliang Li,Fan Hong,Haotian Zhou,Jun Wang,Yu Xu,Hongliang Ren,Cheng Fei,Yuegang Tan,Zude Zhou
标识
DOI:10.1177/02783649251328373
摘要
This paper proposed a step-coated Fiber Bragg Grating (FBG) 6-axis Force/Torque (F/T) sensor with self-fault tolerance for robot-assisted surgery interaction state monitoring. Step-coated FBG achieves a twin reflection spectrum of double-wavelength peaks and hardware redundancy for a single branch, leading to temperature self-compensation without additional FBG and supporting enough data for fault tolerance. Eight step-coated FBGs have been cross-tensioned and welded into the runway-shaped beam from the designed sensor to raise long-term and high temperature (200°C) stability. The regulating mechanism of the step-coated FBG’s temperature and strain sensitivity by coating size was established, providing a modified theory of FBG sensing. The designed sensor’s performance characterization and optimization model has been derived considering the parameters of coating and runway-shaped beam, leading to a low max full-scale error (MFSE, 3.46%) and inter-dimensional coupling error (8.96%). The step-coated FBG and least squares matrix libraries are united to form a hardware and algorithm collaborative fault-tolerant strategy with temperature self-compensation, achieving a real-time (0.2 ms) recovery rate of 25.7%. Experiment results demonstrate that the temperature self-compensation and long-time stability measure errors are less than 4.39% and 5.93%, respectively. A robot-assisted surgery system equipped with the designed sensor and Multi-channel Convolutional Gated Recurrent Unit-based (MCGRU) identification model has been constructed to monitor the force and stages during orthopedic procedures. The MCGRU realized a stage identification accuracy of 97.7%, only relying on wavelength data. The breakthrough detection and force control experiments have been implemented in bovine bone, animal skull, and corpse skull to further verify the design sensor feasibility and effectiveness for robot-assisted orthopedic surgery.
科研通智能强力驱动
Strongly Powered by AbleSci AI