Long cable-driven continuum robots (CDCRs) often suffer from backlash due to cable elongation and friction, which severely impairs control accuracy. This paper presents a novel in-situ cable displacement sensing approach based on optical flow sensors and an adaptive method for real-time backlash estimation and compensation. These lightweight, low-cost sensors are embedded in the robot’s distal section. Experimental validation was conducted on single-cable setups and complete CDCR configurations, including a straight configuration and a simulated aircraft engine combustor environment with complex passive section geometry. The results demonstrate that the proposed method effectively compensates for the backlash in real time, significantly improving the trajectory tracking accuracy. Under different passive section shapes, the open-loop trajectory tracking RMSE is reduced by up to 66.0% compared to the baseline case with no compensation. The proposed method does not require external illumination or complex computation and exhibits strong robustness to ambient lighting variations, making it suitable for deployment in confined, dark, or flashing environments. This work offers a scalable and practical solution for mitigating backlash in long and flexible CDCRs, thereby enhancing control performance in industrial applications such as combustion chamber inspection.
原文链接
分享
引用
