ABSTRACT In this paper, to enhance the control precision and robustness of continuum robots against noise‐induced joint drift and positioning errors encountered in complex environments, a noise‐tolerant zeroing neural network (NTZNN) based control method is proposed. Starting from the kinematic characteristics of continuum robots, a spatial kinematics model is established, formulating the joint‐space‐to‐task‐space mapping. Then, a second‐order error dynamics formulation incorporating integral compensation is constructed. Leveraging the second‐order formulation, the NTZNN model is developed to derive the joint velocity control law. Furthermore, to mitigate computational complexity and address the singularity problem, an auxiliary matrix is introduced for efficient inverse estimation in the control law implementation. Theoretical analyses conclusively verify the feasibility of the NTZNN‐based control scheme and its robust performance characteristics for high‐precision control of continuum robots. Finally, the superior performance of the proposed control scheme in trajectory tracking accuracy, noise suppression, and real‐time performance is verified through simulation experiments of a three‐link continuum robot manipulator.
Haochen Tang (Wed,) studied this question.