Hybrid Data-Driven Optimization Design of a Layered Six-Dimensional FBG Force/Moment Sensor With Gravity Self-Compensation for Orthopedic Surgery Robot

Abstract

This paper developed a layered six-dimensional FBG force sensor to detect the interaction force between drill and tissue in robot-assisted bone drilling. Eight unique C-shaped beams are arranged in layers within the structure to constitute a force-sensitive 3D printed flexure-body and eight FBGs have been tensely suspended on the concave side of beams, leading to low chirping risk and low cost of the designed sensor. A mathematical sensing model of the designed sensor has been derived by response surface methodology (RSM) with the hybrid data obtained from experiments and the FEM. Multi-objective optimization model of the sensor with consideration of machining deviation has been built, and its measurement isotropy for force and moment has increased by 31.4% and 30.7%, respectively, as well as the enhancement of the capabilities of the inter-dimensional decoupling and toleration for machining deviation. Experimental results demonstrate a low coupling error and a high resonant frequency of more than 232 Hz. The relative error of the sensor is less than 8.27%. A robot-assisted bone drilling experiment has been implemented to further confirm the feasibility and reliability of the sensor. By innovatively integrating a gravity self-compensation method, the sensor's detected force agrees with the reference result with a less error of 7.4%.