Shape Reconstruction for Flexible Robot Using FBG Sensor

This is the authors' abstract. We don't add key points for this paper.

Flexible robots have shown great potential in medical applications. Highly accurate shape perception methods for flexible robots can effectively enhance their operability in complex environments. In this paper, an improved 3D shape reconstruction algorithm based on the Frenet-Serret formulas is proposed to reconstruct the robot curve. The complex 3D curve shape can be reconstructed by segmenting and stitching the curvature and torsion information collected by fiber Bragg grating sensor. The calibration experiments prove the reconstruction accuracy and calculation speed of the algorithm. The shape sensor for the flexible robot is fabricated, 2D and 3D shape sensing experiments are conducted. The average Euclidean errors of the reconstructed curve endpoints are 1.120 mm and 1.358 mm, and the average Frey interval distances are 1.922 mm and 2.533 mm, respectively. The experimental results showed that the designed shape sensor has high accuracy to meet the actual sensing requirements.