Cable-driven continuum surgical robots face significant challenges in nonlinear physical modeling, parameter uncertainty, hysteresis and external disturbances. Achieving high-precision trajectory tracking to assist physicians in natural orifice transluminal endoscopic surgery remains highly demanding. This paper aims to systematically review and summarize research on trajectory tracking control in this field to provide references for future studies.
A systematic search of Scopus and ScienceDirect was conducted, combined with keyword co-occurrence and clustering via VOSviewer. Control methods are classified into model-based and model-free approaches.
The model-based control methods are divided into kinematics-based and dynamics-based approaches. Kinematics-based methods, including Jacobian inverse, advanced and static control, suit low-speed and low-precision tasks. Dynamics-based methods, categorized as parameter-determined or parameter-uncertain depending on known system parameters, are better suited for complex surgical environments requiring high-precision dynamic performance. Model-free methods rely on data-driven techniques, such as neural networks and reinforcement learning, to compensate for hysteresis, friction and nonlinear disturbances in cable-driven systems, making them suitable for environments where accurate models are hard to establish.
The paper classifies major trajectory tracking methods, clarifies their applications and proposes future directions in design, control, artificial intelligence and multimodal perception to advance intelligent and clinical use of continuum surgical robots.
