.png){kind=icon}
.png){kind=icon}
Return
Adaptive Load-Dependent Sim2Real Framework for Path Tracking Toward Tendon-Driven Continuum Robots
.png){kind=icon}
PRE
AI
DOI：10.1109/TMECH.2025.3566053.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Continuum robots (CRs) are crucial for executing intricate tasks in constrained anatomical spaces. However, challenges such as actuator latency, backlash, and tendon pretensioning can limit their effectiveness. In this article, we introduce an adaptive load-dependent simulation to the real-world framework for path tracking in tendon-driven CRs, essential for enhancing the dexterity and precision required in minimally invasive surgical procedures. We propose a fixed-pose inverse kinematics control method integrating virtual degrees of freedom into end-effector planning to facilitate six-dof navigation. Then, it establishes a relationship between overall tendon loading and motor actuation, paving the way for employing the learning-based neural network for accurate calibration of the robot's physical model, enabling arbitrary trajectory tracking without the need for external shape-sensing sensors. Our method maintains a maximum 3-D root-mean-square error (RMSE) of 5.01 mm across diverse trajectory types, achieving a directional discrepancy RMSE of 0.114 radians. This demonstrates a significant improvement in tracking accuracy, reducing tracking errors by 86.1% compared to the noncompensated approach. The findings of this study highlight the potential of our approach to bridge the gap between simulation and real-world applications, ultimately enhancing operational efficiency and patient outcomes in robotic-assisted surgeries.
Keywords:
Continuum robot (CR)
hypothesis compensation
path tracking
Journal
I
IF:
0
Papers:
112
Citations:
0