A Synchronous-Optimized and Safety-Improved Framework for Human-Robot Interaction in Robot-Assisted Knee Arthroplasty

Abstract

Shared control is the most commonly used physical human-robot interaction (pHRI) in robot-assisted orthopedic surgery. However, challenges persist in tasks such as robotassisted knee arthroplasty, particularly in terms of motion delays and inadequate three-dimensional (3D) constraints. In this letter, a synchronous-optimized and safety-improved pHRI control framework is presented to perform coordinated operation in complex constrained task spaces. A time-series-based deep learning model is employed to predict surgeon's intentions and anticipated trajectories, improving synchronization performance. Then, a dual-layer fuzzy adaptive constraint strategy is developed to dynamically adjust the constraint scope, ensuring precision, flexibility, and safety within the 3D task space. Experimental results indicated that, in comparison to conventional adaptive admittance control, the proposed method can increase operational compliance by 27.56% and 12.84% during translation and rotation, and reduce interaction forces by 31.88% and 23.76% during translation and rotation, respectively. Additionally, the proposed method demonstrated a mean boundary control accuracy of 0.479 mm, representing a 13.07% improvement over the commonly used forbidden-region virtual fixtures method.